Giessener Beiträge zur Fremdsprachendidaktik Miriam Neigert New Perspectives on Older Language Learners A Mixed Methods Study on the Temporal Self of Young-Old EFL-Learners in Germany New Perspectives on Older Language Learners GIESSENER BEITRÄGE ZUR FREMDSPRACHENDIDAKTIK Herausgegeben von Eva Burwitz-Melzer, Wolfgang Hallet, Jürgen Kurtz, Michael Legutke, Hélène Martinez, Franz-Joseph Meißner und Dietmar Rösler Begründet von Lothar Bredella, Herbert Christ und Hans-Eberhard Piepho Miriam Neigert New Perspectives on Older Language Learners A Mixed Methods Study on the Temporal Self of Young-Old EFL-Learners in Germany © 2019 · Narr Francke Attempto Verlag GmbH + Co. KG Dischingerweg 5 · D-72070 Tübingen Das Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes ist ohne Zustimmung des Verlages unzulässig und strafbar. Das gilt insbesondere für Vervielfältigungen, Übersetzungen, Mikroverfilmungen und die Einspeicherung und Verarbeitung in elektronischen Systemen. Internet: www.narr.de eMail: info@narr.de CPI books GmbH, Leck ISSN 0175-7776 ISBN 978-3-8233-8311-6 (Print) ISBN 978-3-8233-9311-5 (ePDF) ISBN 978-3-8233-0163-9 (ePub) Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http: / / dnb.dnb.de abrufbar. 9 10 11 12 1. 15 1.1 16 1.2 17 2. 21 2.1 21 2.1.1 23 2.1.2 24 2.1.3 26 2.1.4 31 2.2 32 3. 35 3.1 35 3.2 37 3.2.1 38 3.2.2 39 3.2.3 42 3.2.4 47 3.2.5 51 3.3 54 3.3.1 55 Inhalt Statement of Academic Conduct . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . List of Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Elephant in the Classroom - Demographic Changes, Adult Education, and Language Learning . . . . . . . . . . . . . . . . . Chapter Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Who are the Young-Old? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Towards a Definition of the Young-Old . . . . . . . . . . . . . . . . . . . Biological Dimension . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chronological Dimension . . . . . . . . . . . . . . . . . . . . . . . . . Sociological Dimension . . . . . . . . . . . . . . . . . . . . . . . . . . Psychological Dimension . . . . . . . . . . . . . . . . . . . . . . . . . The Young-Old Language Learner . . . . . . . . . . . . . . . . . . . . . . . Self-Concept, Its Temporal Facets, and Foreign Language Learners . . . . . A Note on Self-Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . Self-Concept is… . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Back to Basics: Concepts . . . . . . . . . . . . . . . . . . . . . . . . . Of Pyramids, Jellyfish, and Molecules: Situating the Self and Self-Concepts in our Minds . . . . . . . . . . . . . . . . . . . Self-Concept Formation . . . . . . . . . . . . . . . . . . . . . . . . . . Definition by Demarcation: What Self-Concept is Not Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Temporal Self . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Possible Selves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.2 58 3.3.3 59 3.4 60 3.4.1 61 3.4.2 65 3.5 66 4. 69 4.1 70 4.1.1 72 4.1.2 75 4.1.3 78 4.1.4 79 4.2 84 4.3 87 4.3.1 89 4.4 96 5. 99 5.1 101 5.2 108 5.3 114 5.4 116 5.4.1 116 5.4.2 121 5.4.3 124 5.4.4 133 5.5 135 6. 137 6.1 138 6.1.1 139 Self-Determination Theory (SDT) . . . . . . . . . . . . . . . . . . Self-Discrepancy Theory . . . . . . . . . . . . . . . . . . . . . . . . . Self-Concept in SLA Research: the L2 Motivational Self-System (L2MSS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . L2MSS: Roots and Application . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methodology, Research Design, and Research Context . . . . . . . . . . . . . . Mixed Methods Research - Why? , What? , How? . . . . . . . . . . . Intermezzo: Research Questions Revisited . . . . . . . . . . . Why? - Reasons for Employing MMR . . . . . . . . . . . . . . What? - Research Instruments . . . . . . . . . . . . . . . . . . . . How? - Finding the Right Mix . . . . . . . . . . . . . . . . . . . . Agree to Disagree? Inference Quality and Handling Divergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Research Context: Adult Education Centres in Germany . . . . Volkshochschule - The Institution . . . . . . . . . . . . . . . . . Summary and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Numbers: The Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . From Theory to Instrument: Development of Questionnaire Items . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Questionnaire Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Administration: Sampling and Ethics . . . . . . . . . . . . . . . . . . . . . Processing: Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Descriptive Statistics: Demographics and Motives of Respondents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inferential Statistics: Multi-Item Likert Scales . . . . . . . . Inferential Statistics: Analysis of Adjective Scales . . . . Discussion in Light of the Hypotheses . . . . . . . . . . . . . . Summary and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Words: The Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Controlled Flexibility? Approaching a Semi-Structured Interview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Overall Structure of the Interview Guide . . . . . . . . . . . . 6 Inhalt 6.1.2 140 6.1.3 141 6.1.4 143 6.2 149 6.3 152 6.3.1 153 6.3.2 154 6.4 157 6.4.1 157 6.4.2 161 6.5 198 7. 202 7.1 205 7.1.1 205 7.1.2 207 7.1.3 209 7.1.4 211 7.1.5 213 7.1.6 215 7.1.7 216 7.1.8 218 7.1.9 220 7.2 221 8. 226 8.1 227 8.1.1 227 8.1.2 231 8.2 232 236 255 259 The Language Learning Biography . . . . . . . . . . . . . . . . . The Pilot Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Re-Connecting to Theory: The Interview Questions . . The Participants: Sampling, Ethics, and Data Collection . . . . . QUAL Data Analysis: Principles and Procedures . . . . . . . . . . . Preparing the Data for Analysis: Transcription . . . . . . Category Formation: Organizing and Interpreting Data Results of the Interview Study . . . . . . . . . . . . . . . . . . . . . . . . . . The Category System . . . . . . . . . . . . . . . . . . . . . . . . . . . . Presentation Sampling: L2-Vision in Focus, L2-Vision as Focus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Numbers and Words: Merging QUAN and QUAL Strands . . . . . . . . . . . . . QUAN-QUAL Integration of Micro-Level Results . . . . . . . . . . Ambition and Intended Effort . . . . . . . . . . . . . . . . . . . . . Resilience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Critical Incidents Abroad and L2-Vision . . . . . . . . . . . . Confidence: Taking it Easy . . . . . . . . . . . . . . . . . . . . . . . . Defying the Effects of Ageing . . . . . . . . . . . . . . . . . . . . . Now and Then: Ways of Learning a Language . . . . . . . Now and Then: Learning Context . . . . . . . . . . . . . . . . . . Perceived Language Learning Progress . . . . . . . . . . . . . Language Learning and the Social Sphere . . . . . . . . . . . Discussion in View of Research Questions . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . So, What? - Implications for the Foreign Language Classroom Implications for Instructors . . . . . . . . . . . . . . . . . . . . . . . Implications for the Institutional Level: The vhs Centres Beyond this Study: New Questions, Further Research . . . . . . . Works Cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix A: Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix B: Letters to Heads of Language Departments . . . . . . . . . . . . . . . . . . 7 Inhalt 260 261 265 266 269 270 277 278 Appendix C: Letter to vhs Centre Language Instructors . . . . . . . . . . . . . . . . . . . . Appendix D: Interview Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix E: Informed Consent - Interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix F: Sample Learning Biographies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix G: Transcription Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix H: Interview Summaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Abbildungsverzeichnis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tabellenverzeichnis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Inhalt Statement of Academic Conduct I certify that the research described in this dissertation has not already been submitted for any other degree. I certify that to the best of my knowledge all sources used and any help received in the preparation of this dissertation have been acknowledged. For this research project, ethics approval has been obtained by the Hessischer Volkshochschulverband (hvv), Frankfurt a. M., Germany, and Justus-Liebig University, Gießen, Germany. It has been approved by the Faculty of Arts Human Research Ethics Committee, at Macquarie University, Sydney. Ich erkläre: Ich habe die vorgelegte Dissertation selbständig und nur mit den Hilfen angefertigt, die ich in der Dissertation angegeben habe. Alle Textstellen, die wörtlich oder sinngemäß aus veröffentlichten oder nicht veröffentlichten Schriften entnommen sind, und alle Angaben, die auf mündlichen Auskünften beruhen, sind als solche kenntlich gemacht. Sydney, 3 rd June 2017 Miriam Neigert Abstract Despite major demographic changes in recent years and in the future, a uniform view of adult language learners has prevailed in foreign language research for a long time. Young-old learners - the participants of my study - are an age group, which spans from approximately 60 to 80 years of age and is in a tran‐ sitional phase from work-life to retirement (and beyond that time) and still shows a high interest in life-long learning. My thesis investigates how young-old learners perceive themselves as foreign language learners, by taking a closer look at their L2 self-concept and its temporal facets, i.e. their past, actual, future/ ideal L2 selves. To do so, my research brings together concepts from gerontology, psychology, adult education, and foreign/ second language research. I have utilized a mixed method approach for my research to gain a better understanding of the complexities of language learner self. Thus, my study combines and analyses a quantitative data strand (a survey study with 195 res‐ pondents learning English at German adult education centres; German: Volks‐ hochschule) and a qualitative data strand (21 in-depth interviews with young-old English learners at a local adult education centre). The findings of the study indicate that young-old language learners’ priorities with regard to foreign language learning undergo a shift with old age and retirement. It is a shift to‐ wards an increased value of the social sphere in a language class, as well as the intention of defying the effects of ageing by learning a foreign language (and with it practicing the long-advocated ‘lifelong learning’). Moreover, the study illustrates the importance of critical incidents abroad in the past as well as a connection to an L2-community when it comes to shaping young-old language learners’ degree of resilience and quality of an L2-vision (Dörnyei 2014). With this, my thesis offers a new, more differentiated perspective on older language learners. Its distinct contribution lies in uncovering the importance of resear‐ ching the temporal facets of their language learner self. List of Abbreviations CEFR Common European Framework of Reference for Languages EFL English as a Foreign Language ESL English as a Second Language FL Foreign Language FLC Foreign Language Classroom FLL Foreign Language Learning hvv Hessischer Volkshochschulverband (Hessian Adult Education Associa‐ tion) L2 Second/ Foreign Language MMR Mixed Methods Research L2MSS L2 Motivational Self-System QUAN Quantitative Data Strand QUAL Qualitative Data Strand RQ Research Question SLA Second Language Acquisition SDT Self-Determination Theory vhs Volkshochschule (adult education centre; vhs centre) Acknowledgements This thesis is a product of several years of study and research. In those years, many people have supported and inspired me. In a way, they also made this thesis what it is. At this point, I would like to express my gratitude to them: First and foremost, without the excellent academic support of my supervisors Prof. Dr. Michael Legutke and Prof. Dr. Martina Möllering this thesis would not have come to fruition. Despite having written hundreds of pages for this thesis already, it is hard for me to find the right words to can express my gratitude to them. I consider myself very lucky that I have been under their guidance while undertaking this PhD-project. With their openness to my research project, end‐ less patience, and incredibly valuable feedback, they pushed me to new levels and, in this way, helped me grow as a person and researcher. I wish to thank Martina and Mitch for giving me the chance to experience both German and Australian academic worlds through their joint supervision - I am especially grateful for the funding and research support through the International Mac‐ quarie Research Excellence Scholarship (IMQRES) scheme. Moreover, I also would like to express my gratitude to Prof. Dr. Jürgen Kurtz, who was willing to act as my associate examiner at Justus-Liebig University, and who has been very supportive especially in the early stages of my candidature in Gießen. I thank the study participants and my former course participants for their time, their trust, and cooperation. All of them contributed significantly to this research project by sharing their experiences with me. I have learnt so much from them during my time as a teacher and later as a researcher. I also greatly appreciate the support I received from other teachers at participating vhs centres in the state of Hesse and the coordinating heads of departments at local vhs centres and the central Hessian adult education association hvv. In particular, I would like to thank Werner Leipold and Barbara von der Meden for their support. Many thanks also go out to the fellow researchers and professors in the GCSC Research Colloquium at Justus-Liebig University, because I learned so much through the insightful discussions in our meetings and from their input. I also benefited greatly from my HDR writing group which was initiated by Dr. Mi‐ chelle Jamieson at Macquarie University. I would like to thank my fellow-writers in this group - Yujie Zhang, Lina Qian, Suzanne Eiszele-Evans, Marisa Della Gatta, and Anthony Andrist - for their valuable feedback on my writing, as well as for giving me an opportunity to learn from their writing processes, as well. Moreover, I would like to thank Prof. Dr. Naren Chitty for giving me the op‐ portunity to present my research progress at the International Graduate Con‐ ference at Tsinghua University, Beijing, in 2015 and at Macquarie University, in 2016, as it helped me to further refine my research design. Furthermore, I would like to express my deepest gratitude to some of my closest friends who I have known for so long and who greatly supported me all the way through highs and lows: Hanna Reinshagen, Annika Schäfer, Leila Schmidt, Sabrine Tryba, and Rose Fitz-Patrick. I wish to thank them for keeping me grounded, making me laugh, sharing a cup of coffee or glass of wine when needed, and always giving me a new perspective when things seemed difficult! Furthermore, I was lucky to have fellow researchers, colleagues, friends, and family in Germany and Australia, who inspired and encouraged me in so many ways throughout my PhD-candidature: Susanne Bauer, Reinhard Bentrup, Dr. Susanne Binder, Franziska Burghardt, Dr. Bruce Doar, Constanze Dreßler, Dr. Ulrike Garde, Dr. Ursula Ibaraki, Esther Li, Nicole Mutlow, Susana Catalina Prat, Birgit Sauer, Eva Schmidt, Inke Schmidt, Tamara Seher, Dr. Ian Stephen, Dr. Brangwen Stone, Sebastian Stuhlmann, Juliane Witzenberger, and Dr. Marja Zi‐ belius. Thank you! The above expressions of gratitude show that undertaking this PhD has been an exciting and rewarding journey. Finally, I would like to thank my parents, Michael and Marlies, who have been at my side during all my journeys in life so far, and who were always there for me when I needed them. 13 Acknowledgements 1 All participants’ names in this study have been changed. 1. Introduction The older I get, the braver I become. Jennifer, 68, retired. Jennifer 1 had only recently joined my English class at a local adult education centre (Volkshochschule, hereafter: vhs centre). In her interview, she told me about her doubts before she joined this course, which was specifically designed to practice conversation. But despite her doubts, she gave it a try and did not regret her choice. Yet, her statement indicates that as a younger woman, she possibly would have refrained from doing so. It is learners like Jennifer who made me want to find out more about how older learners approach learning a language. On my search for answers, which started long before my PhD during my studies at university, I discovered that research on how foreign or second languages are learned and taught has widely neglected older adults (see chapters 3 and 4). Very often research in this field is based on the assumption that adul‐ thood (i.e. everything past the age of 18) is a stable and uniform concept (but see: Kade 2009; Dannefer & Settersten 2010; Schmidt-Hertha 2014). Moreover, as soon as age and ageing are explored in conjunction with this research, it is explored in terms of the possibilities and boundaries of second language acqui‐ sition (e.g. critical period hypothesis; Singleton & Ryan 2004; Schramm 2010; Herschensohn 2007). Therefore, this thesis aims to provide new perspectives on older language learners by studying their language learner self and its temporal facets: how do older learners perceive their past, current (or actual), and future language learner selves? In order to move away from a uniform view of adul‐ thood, my study will narrow down the focus on a specific group of older learners I refer to as ‘young-old’ (see chapter 2). Young-old learners are between 60 and 80 years of age, mostly retired, and are still active in that they are interested in starting or continuing learning English as a foreign language. It is hoped that my study can open up new ways to view older adults who learn foreign lang‐ uages and that it initiates further discussions about adapting language learning programs to their expectations and needs. 1.1 The Elephant in the Classroom - Demographic Changes, Adult Education, and Language Learning Industrialized nations, such as Germany in the context of this study, have wit‐ nessed a demographic shift leading to an ageing society. More and more people grow older and remain relatively healthy and active for longer, while at the same time birth rates remain at a low level. This has its impacts on the structure of our life course and also on the economy in that people will be able and possibly required to work longer throughout their lives. Simultaneously, “perforated” working lives have become more common in industrialized countries resulting in more frequent professional changes. These societal and economic shifts re‐ quire people to adapt to new circumstances and ultimately to be willing to keep learning throughout their lives - the keyword here being “lifelong learning” ( Jarvis 2009a; Komp & Aartsen 2013). Yet, while this has become a widely debated issue in politics and the economy, it appears that the role of age and ageing in educational research is only slowly being recognized as a pressing issue (Hammond 2005; Jovic & McMullin 2011; Kolland & Klingenberg 2011: 19; Käpplinger 2014). In other words, older lang‐ uage learners and their needs have been the ‘elephant in the classroom’, which will be addressed in the course of this study (Ramírez-Gómez 2016; Berndt 2003; Doff 2005; Raasch 2005). If we are to address the growing demand for lifelong learning, researchers and educational practitioners need a better understanding of learning in adulthood, and, more specifically, in later adulthood. As mentioned above, research - including foreign language research - has treated adulthood as a uniform construct and for the most part conducted empirical studies with younger adults (see chapter 3). However, it does not take long for teachers to notice differences between a language learner who is 28 years of age and a language learner who is 68 years of age and retired - both being adult learners. Thus, I hope to redirect the focus of foreign language research to older language learners and add to a growing field of “foreign language geragogy” (Berndt 2003; Ramírez-Gómez 2016). With this, my study contributes to the research on life‐ long learning by looking at people who - quite literally - have chosen to keep learning their whole life, even beyond their working life. The aforementioned vhs centres (see chapter 4) are the hub for lifelong learning in Germany as they are the main - and in some areas the only - provider of continuing education. Vhs centres in Germany attract many of the older learners who are interested in starting to learn or brushing up on their English. This makes this learning context a suitable starting point for carrying out my research on how the young-old learn languages. 16 1. Introduction To better understand older learners, I decided to investigate their language learner self-concept and its temporal facets (Dörnyei 2005; Mercer 2011; Mercer & Williams 2014). As this study will exemplify, using a mixed method research design which combines quantitative and qualitative data enables researchers in the field to explore the complexities of language learner self (see chapters 4 to 7). It is hoped that the results of this study will inspire other foreign language researchers to pursue this direction of research. For institutions such as the vhs centres in Germany, it is important for them to find out how to attract and retain older learners to their course programmes. This study provides a foundation for future decision-making by these institutions. My research aims to provide in‐ structors who are working with older learners with a better understanding of why and how young-old language learners approach learning English inand outside the classroom. 1.2 Chapter Outline To achieve a better understanding of young-old language learners, the following chapters cover the theoretical overviews (chapters 2 and 3), the methodological framework and research context for this research project (chapter 4), and finally the empirical work conducted (chapters 5, 6, and 7). A concluding chapter briefly outlines the results and discusses the implications of the research outcomes. With regard to the theoretical foundations, chapter 2 will provide a better un‐ derstanding of the participants in this study: the young-old language learners. What differentiates the young-old learners from the participants in most other research studies on language learning is their age. As chapter 2 will highlight, there is more to age and ageing than the mere chronological perspective. Using different dimensions of ageing - biological, chronological, sociological, and psychological age - chapter 2 provides a working definition of “young-old” for this study. The chapter ends with a brief overview of previous studies on older language learners in the field of SLA, which reveals the need for more research in this area. With a focus on self-concept within the realm of foreign and second language research, chapter 3 addresses a second theoretical foundation of this study - self-concept within psychological research. This detailed overview includes the nature of self-concept and its formation over the life course. This is followed by a discussion of other self-terms, which are also commonly investigated in SLA research, to further clarify the meaning of self-concept. The focus is then further narrowed down to the temporal aspects of self-concept and theories connected 17 1.2 Chapter Outline to this: possible selves, self-determination theory, and self-discrepancy theory. What are the dynamics between our past, present, and future self-concepts? What makes us work towards or away from a certain future self ? Do we learn a foreign language because of our personal ideals and aspirations, or because we think we are expected to do so? In its final sections, chapter 3 connects these temporal aspects of self-concept to recent SLA research on the language learner self. The chapter introduces the L2 Motivational Self-System (Dörnyei 2005, 2009) which comprises ideal as well as ought-to L2 self guides, and the L2 lear‐ ning experience. It is hoped that this model helps to provide a better under‐ standing of the young-old language learners in this study. Having presented and discussed the theoretical foundations, chapter 3 concludes with the major qu‐ estions addressed by this study. Chapter 4 outlines the study’s research design, in order to find answers to the research questions. A mixed-method design has been chosen to investigate the young-old language learners’ L2 self-concepts and their temporal facets. Com‐ bining the results of a survey (quantitative data strand; QUAN) and an interview study (qualitative data strand, QUAL) can provide an expanded understanding of the complexities of language learner selves. Chapter 4 also introduces the research context: adult education centres - so-called Volkshochschulen (vhs centres) - in Germany to which young-old language learners typically turn if they intend to learn or brush up on a foreign language. Chapters 5 and 6 present how the theoretical concepts introduced in earlier chapters were operationalized in the quantitative and qualitative instruments and how the respective data were collected and analyzed. Chapter 5 outlines the development and analysis of the questionnaire on the self-concept of young-old language learners. It brings together the results of 195 respondents, who are learning English as a foreign language at vhs centres in the German state of Hesse, and correlates the results with factors relevant to the research interests. The qualitative data strand evolves around interviews with 21 young-old lear‐ ners at a local vhs centre. Similar to the survey, the interview guide aimed elicit information about young-old language learners. These two chapters - 5 and 6 - also present and contextualize the research study’s results in view of the re‐ search questions and the following merging stage. Following this, chapter 7 merges the two data strands presented in chapters 5 and 6 and draws meta-in‐ ferences based on the points of convergence and divergence between quantita‐ tive and qualitative perspectives. The conclusion in chapter 8 gives a brief overview of the results and examines the implications of the results of the study for adult education institutions such as vhs centres in Germany and foreign language instructors, who are teaching 18 1. Introduction older learners. This final chapter also presents directions for further research that emerged in the course of the research project. 19 1.2 Chapter Outline 2. Who are the Young-Old? If, as seems presently true, the young-old will not form a strong age-group identification of their own, they might become the major agents of social change in building the age-ir‐ relevant society. If they create an attractive image of aging, thus allay the fears of the young about growing old, and if they help to eradicate those age norms which are cur‐ rently meaningless and those age attitudes which are currently divisive, they will do the society an untold service. Theirs is an enormous potential. - Neugarten, 1974: 198 - Ageing is something all humans have in common from the day they are born. This makes age a crucial part of our identities. In this chapter, however, we will see that there is more to age than the number of candles on our birthday cakes, that is, our chronological age. In the following, it will become clear in what ways age and ageing are multi-dimensional and constructed. The overall aim of this chapter is to give a clearer picture of the young-old and their age-specific cha‐ racteristics. Referring to Neugarten’s quote above, the age group ‘young-old’, which has been largely overlooked in language teaching and learning research, is still widely exposed to persisting age-bias. How this affects their language learning self-concept and in what ways instructors can foster their language learning shall be at the core of this study. 2.1 Towards a Definition of the Young-Old Internal and external factors affect how we age and they add to the individuality of the ageing process. Some of these factors we can control more easily than others. It is, for example, mostly our choice if we live healthily without smoking and follow a good diet, hoping to decrease our chances of serious illnesses as we get older. But some other factors, which impact how our bodies and minds age, we cannot influence. For example, current cultural notions of age and ageing influence our perception of what is considered old or how we evaluate old age. Furthermore, political or historical events, such as wars or crises, impact the lives and ageing of a cohort or generation. It is for reasons like these that variability among individuals increases with age (Dannefer & Settersten 2010: 4). This, in turn makes it difficult to come to a definition of an older age group such as the young-old. I have taken the term young-old from literature on lifespan development and gerontology - the most prominent use of it being in an article by Neugarten (1974). In the 20 th century, life expectancy (also referred to as population ageing or societal ageing) has been increasing rapidly and researchers have started to further divide the old age phase of life (Phillips et al. 2010: 171; Nolda 2008). The former one-dimensional old age phase (after retirement) is now more commonly divided into young-old (Third Age) as well as old-old and oldest old (fourth age) (Laslett 1987; Neugarten 1974, Bromley 1990: 42). Alongside the term young-old, the term ‘third agers’ has been increasingly used when referring to members of an age group spanning from shortly before, to approximately 10 or 15 years after, retirement (see section 2.1.3). A perspective I would like to use as a starting point for my discussion of the young-old appears in Neugarten’s (1974) article on age groups in American so‐ ciety. According to Neugarten, perceptions of the life cycle and stages in life have changed. Over the last centuries, concepts such as childhood or middle age have been introduced or become more explicit in our minds due to societal changes (e.g. industrialization). She postulates that the 20 th century has been the onset of forming another distinctive age group or phase in life: the young-old and their demarcation from the old-old (1974: 190 f.). She defines the young-old as “relatively healthy, relatively affluent, relatively free from traditional res‐ ponsibilities of work and family and who are increasingly well educated and politically active” (1974: 187). Even though Neugarten acknowledges the limited meaning of chronological age for defining an age group such as the young-old, she sets the age range at 55 to 75. Nevertheless, she sees a more important or defining characteristic of the young-old in the sociological marker of retirement (1974: 191). From these initial words on the young-old, it becomes clear that given the increased variability in old age as well as the attempt to set boundaries in terms such as chronological age, the definition of ‘young-old’ sails between vagueness and oversimplification. We find four different dimensions of age or ageing in the gerontological li‐ terature (Karl 2009: 23; Phillips et al. 2010: 12): biological, chronological, socio‐ logical, and psychological. Departing from Neugarten’s definition here, I will use these four different dimensions and discuss how the young-old are positi‐ oned within these dimensions to find a working definition of the young-old for my study. I will discuss all of the dimensions in the following. Due to the nature of my research project, however, I will focus more strongly on the sociological and psychological dimensions. 22 2. Who are the Young-Old? 1 Similar to my argumentation, Smith and Ryan (2016: 305) use the ability to participate in a research study as an indicator for vitality in their oldest-old study participants (aged 85+). 2.1.1 Biological Dimension The biological dimension of age and ageing - also referred to as senescence - is about how our body or physiological system ages. The ageing process of our bodies is dependent on our genetic disposition, our behaviour (e.g. diet and exercise), and to some extent also our environment (e.g. exposure to sun or chemicals). The physiological functions subject to change represent the so-called biomarkers and help to indicate the biological ageing process. These factors make our biological ageing a fairly unpredictable process: while one person needs a hearing aid at age 50, another person’s ears function very well until they are 75 (Rowland 2012: 167). Thus, this dimension alone is not very useful for describing the young-old learners in my study, as it is hardly possible to universally define biomarkers that would negatively impact their language learning process. Neugarten in her definition above describes the young-old as relatively healthy and active (see also Bromley 1990: 45). The word ‘relatively’ again shows how difficult it is to draw a clear picture of this age group. What, for example, would the word ‘relatively’ entail for the young-old learners in my study? A broad definition of relatively healthy in the domain of foreign language learning would include language learners still independent enough to be able or to be enabled to undertake language learning away from their homes. 1 Generally this would also mean that sensory impairments - visual and auditory - as well as cognitive impairment (even though specific language classes for participants with a mild dementia exist) have not progressed to a state making it too hard for young-old learners to achieve their language learning goals. Additionally, with constantly improving medical treatment and intervention programmes (e.g. improved hearing aids or laser treatment) the biological dimension has become less of a characteristic effectively differentiating the young-old from, for example, the middle-aged or old-old with regard to language learning (Phil‐ lips et al. 2010: 15 f.). Again, we see how this age dimension presents us with more vagueness or problems when trying to find a working definition of the young-old. 23 2.1 Towards a Definition of the Young-Old 2.1.2 Chronological Dimension We may expect that our successors in the conceptualization of age and the life-course will replace the nominal numerical adjectives and descriptions with which we have been dea‐ ling with concepts of much greater epistemological and practical effectiveness. - Laslett 1994: 445 f. - The chronological dimension of age focuses on the number of years an indivi‐ dual has lived, i.e. it is determined by the year of birth. Every ID or passport thus makes a statement about one’s chronological age. Entering school or retirement, getting a driver’s license, being able to vote or allowed to drink alcohol - all of these things are opened up or restricted by a certain chronological age. They are not universal and they can change, i.e. they are arbitrary, which is why some of these examples are regularly at the centre of heated debates - as in the case of setting a retirement age (Midwinter 2005: 10; Rowland 2012: 167). Like the aforementioned increasing variability in the biological age dimension as indi‐ viduals age, a person’s chronological age often sheds little light on the particu‐ larity of the ageing process (Dannefer & Settersten 2010: 5). This means that sometimes we find more differences between people of a particular cohort than between individuals who are 20 or more years apart (Lehr 2006). Researchers have set the chronological age boundaries or spans for the young-old at different ages. While Neugarten (1974) cautions against only using chronological age, she places the young-old at 55 to 75 years of age. However, she set this age range in the mid-1970s and it is thus questionable today since the age span has shifted upwards due to further life extension or population ageing over the last forty years (Phillips et al. 2010; Victor 2010). Baltes and Smith write about the factors that have played into this development: In concert, better material environments, more advanced medical practice, the im‐ proved economic situation of older persons, more effective educational and media systems, increased psychological resources such as reading, writing and computer literacy, and many other related factors allow older persons to approach their own maximum life span in healthier and more vital conditions. (2003: 126) However, we need to keep in mind that this development of approaching one’s maximum age in a better condition does not affect all young-olds of a cohort equally due to various factors such as their social background or environment (see next section). As we will see in the following sections, today’s young-old, if reduced to their chronological age only, can differ remarkably compared to the young-old 20, 30, or 40 years ago in terms of living longer and of an incre‐ asingly better overall physical and mental state. Aside from Neugarten’s age 24 2. Who are the Young-Old? span for the young-old, we find differing chronological age spans of the young-old in other literature ranging from 10-year spans to 15-year spans and 20-year spans (see table 2.1). Age span (years) Age range (years of age) 10 66-75 (Bromley 1990) 65-74 (Rowland 2012) 15 60-75 (Kade 2007) 20 55-75 (Neugarten 1974; Singleton & Ryan 2004 60-80 (Kruse 2008) Table 2.1 Chronological Age Spans for the Young-Old From these different age ranges we see, firstly, that the ‘core’ of the young-old in terms of chronological age lies between 65 and 71 years of age and the ma‐ ximum boundaries are set at 55 and 80 years of age. Secondly, we can see from these differing age spans for the young-old in the literature that chronological age is a fuzzy concept - at least if it is employed alone - and is not effective for defining members of an age group (cf. Cavanaugh & Blanchard-Fields 2002: 1; Laslett 1987). Thirdly, researchers here gave little or no explanation of how they arrived at the given age span for the young-old. If, however, researchers are justifying the age boundaries, they usually do so by referring to two factors: socio-political and epidemiological. The first factor’s entry point to older age being 60+ or 65+ has been usually grounded on the welfare systems in industrialized Western countries, i.e. the onset of retirement. As mentioned above, retirement age is an arbitrary figure. Nevertheless, as we will see in the next two sections, the factor of retirement is a crucial threshold to what is considered old. The second factor, epidemiology, sets another boun‐ dary at 80+ or 85+ due to an age-related increase in chronic diseases, mortality, and the use of health and social care resources (Victor 2010: 64; see also: Alley & Crimmins 2010). This further supports the age span 60 to 80 as an appropriate boundary for the young-old in this study. In addition to the factors of retirement and epidemiology, the age range 60-80 encompasses a typical 20-year timeframe for a generation. In the case of the current 60 to 80 year-olds, we are dealing with individuals who grew up or were born during or shortly after a period of war, i.e. before 1955. However as dis‐ cussed above, setting a chronological age frame for the young-old group de‐ 25 2.1 Towards a Definition of the Young-Old ceives us into black-and-white thinking. Setting the age range for the young-old group at 60 to 80 in my study is thus more a preliminary guideline before focu‐ sing on more fruitful defining characteristics of the young-old through the so‐ ciological and psychological lens. This also means, for instance, that in the data collection and analysis of this study, 59-year-old or 82-year-old individuals might also be considered young-old due to such factors as the stronger defining influence of the factor of retirement on them or their strong sense of belonging to a particular cohort or generation (e.g. Baby Boomers or Silent Generation). 2.1.3 Sociological Dimension Age-identification is getting increasingly shaky, in good part because social ageing - that is, the roles which society cast us into at different ages - is getting more and more out of step with biological ageing. - Young 1991 in Midwinter 2005: 10 - In the sense of ‘stage over age’ (Midwinter 2005: 11), the sociological dimension is the most essential one for defining my understanding of the young-old in this study. The discussion in the previous sections showed that there is a need to move away from a mere biological and chronological view of age when descri‐ bing members of an age group such as the young-old. These views tend to bear very little differentiation and they have added to a negatively blurred image of old age (e.g. the so-called “deficit model” of ageing: Bromley 1990: 8; Kade 2007: 16 f.; Andrew 2012: 57). Chronological age, for instance, is still part of the defi‐ nition of the young-old by Neugarten and other researchers. While still using the term ‘young-old’ since it captures the young and old facets of this group, I will move away from strongly focusing on the chronological aspect to give this term a stronger sociological meaning. Age can be socially constructed and as a sociological construct it changes over time as well as between cultures. The 20 th century, for instance, has seen an increase in age consciousness, i.e. using (mainly chronological) age to create and justify age norms that shape social structures as well as what we see as ‘age appropriate’ and life trajectories that comply with a ‘social clock’ (Elder et al. 2003: 6; Dannefer & Settersten 2010): at what age should one get married, move out of home, stop driving, stop working, etc.? If we look at past centuries or even millennia, we notice that attitudes to old age have changed in different societies from seeing it as a source of wisdom to regarding it as being a burden on a nation’s economy ( Johnson 2005; Schmidt-Hertha 2014: 18). Connected to this, the time we are born into and the historical events throughout our lives shape how we age and how we perceive ageing. We refer to groups of birth 26 2. Who are the Young-Old? cohorts as particular generations (encompassing a 20to 25-year span), a more prominent example being the Baby Boomer generation born after World War II and the more recent Gen Y (Phillips et al. 2010). We ascribe certain characteristics to each generation affected by the historical change and circumstances in which they grew up (“cohort effects”). For example, we think of the current young-old, namely the young-old in my study, as being born between 1934 and 1954, we are dealing with members of the so-called Silent Generation (early 1920s until early 1940s) and Baby Boomers (late 1940s until mid-1960s). Apart from the effects war and post-war times had on these two generations, they also grew up with an evolving three-box model that shaped the life course in industrialized countries from the mid-20 th century (Hardy 2011; Marshall & Taylor 2005; Dannefer & Settersten 2010: 9 ff.). The three-box model structures life into a pre-employment phase, in which children and adolescents are prepared for the workforce, a second phase of life dominated by becoming part of that workforce, in which one earns and saves for the third and last phase: retirement. Dannefer and Settersten (2010: 10) regard the three-box model as a good example of the institutionalization of the life course since it reflects how “the laws and policies of the state (e.g., compulsory schooling or mandatory retirement) as well as organizations, such as schools (with age-graded classes) and work organizations (with age-graded promotion ranks), define and struc‐ ture the life course.” This life course structure, as Silent Generation members and Baby Boomers have come to live by and expect, is about to disappear due to recent socioeconomic changes such as higher fluctuation in the job market, diminishing public pension funds, and the transition to either higher retirement ages or removing fixed retirement ages completely. Thus, current and future generations can expect to live without this tripartite structure of the life course but rather with a ‘perforated’ life course, i.e. alternating phases of job training, working, and unemployment, while increasingly relying on making provisions for old age themselves (Marshall & Taylor 2005: 575 f.; Dannefer & Settersten 2010: 10; cf. Reutter 2004). Concluding from this, the three-box model shaped one major characteristic of the young-old (in contrast to later generations at least), which is the role of retirement they had expected all their working lives as being a crucial transition into a phase with more freedom to consume and to invest in resources (cognitive, temporal, financial) as they please. Of course, we cannot make rigid connections between the common experi‐ ences of a cohort or generation such as a three-box model of life with every individual young-old’s life path. A major sociological approach to studying age and ageing which takes this into consideration is the life-course perspective (Elder 1974; Elder et al. 2003; Dannefer & Settersten 2010). Dannefer and Set‐ 27 2.1 Towards a Definition of the Young-Old tersten describe how the life course perspective has changed gerontological re‐ search: Old age is no longer viewed as embodying a set of common and universal experiences, nor as a dark period of inevitable decline. Rather, old age is recognized as comprised of a set of experiences that are highly variable across individuals, groups, and nations, and highly contingent on health, wealth, social relationships, social policies, and other factors. (2010: 4) The life-course perspective also employs cohort analysis, which studies the lives of individuals born in a particular year, and this approach “provides an anchor point from which individual trajectories can be constructed and change can be tracked, allowing comparisons across multiple cohorts” (Dannefer & Settersten 2010: 5). Thus, to gain a better understanding of an age group such as the young-old, it is helpful to take a closer look at their ‘social pathways’ (Elder et al. 2003: 8) and identify social, cultural, political, and historical effects influen‐ cing the lives of a birth cohort or even a whole generation. When looking at one generation usually covering 20 years, we need to keep in mind that cohort effects such as the start or end of a war affect individual birth cohorts at different stages of their lives. In the case of the young-old participants in this study who were born between 1934 and 1954, this means that some experienced the end of the World War II as teenagers or young adults and others only as newborns, in utero, or not at all. Early crucial developmental factors, such as malnutrition, educa‐ tion, employment etc. during or after WWII have had different effects on a child, a teenager, or young adult and affected the way individuals or cohorts age. To what extent can we predict that today’s young-old language learners and the way they perceive their learning process have been influenced by such background period factors in their younger years? In the long run, period effects are part of accumulated patterns influencing the ageing of a whole cohort or generation, ‘trickling down’ to the individual. But more individualized accu‐ mulating factors (e.g. the socio-economic status or milieu in which someone grows up) (Elder et al. 2003: 5) also affect the ageing process and need to be considered. Life-course researchers refer to this as ‘cumulative advantage/ disadvantage’ (CAD), which is derived from the ‘Matthew Effect’ (Merton 1968; Settersten & Angel 2011). Alwin (2010: 272) describes this phenomenon as fol‐ lows: “[The S]ocial environment is structured in such a way as to promote the accrual of greater resources to those who already have them - or, cumulative advantage - and the withholding of resources from those who begin with less - or cumulative disadvantage.” Thus, researchers use CAD to explain the intraand inter-cohort variation. It also explains the age stratification from birth to 28 2. Who are the Young-Old? death, i.e. the aforementioned increased variability (and inequality) as we age (O’Rand et al. 2010: 127; Dannefer & Settersten 2010; Settersten & Angel 2011). Applied to the young-old in this study, we can postulate that their early educa‐ tional context during and shortly after WWII was limited or negatively affected which in turn impacted their opportunities to make full use of their cognitive potential. Keeping CAD in mind here again, the impact of this can still vary greatly within a cohort if we take the socioeconomic status or milieu in which a young-old language learner grew up into account. The degree of wealth and socioeconomic stability in times of war and shortly after war in a young-old’s early life gave some individuals of this generation a ‘head start’ (for example regarding their education) and continued to accumulate positively as they age in contrast to other less well-off contemporaries. The degree of education at‐ tained in their younger years, is a valuable predictor for the young-old’s interest in lifelong learning - with the Baby Boomers and Silent Generation displaying higher educational levels than generations before them (Berndt 2003: 14; Phillips et al. 2010: 146 f.). A term that is often used alongside young-old and one that will add to my definition of my study participants is Third Age (Laslett 1987, 1994; Gilleard & Higgs 2002). Laslett’s definition of the Third Age moves away completely from the chronological dimension of age to a more sociological understanding of Third Agers. According to him, Third Age is a phase in life when one has achieved one’s main goals (‘era of personal fulfilment’) and is now free from obligations regarding such things as work or child bearing. He considers the existence of Third Age starting in the second half of the 20 th century in the British population (1987: 137). Before the rise of the Third Age, and reminiscent of the aforementioned three-box model, people moved directly from the Second Age (independence, child bearing, earning and saving for retirement) to the Fourth Age (dependence, decrepitude, death), which still adds a negative connotation to retirement today. With the prolonged life expectancy and quality of life in old age in the 20 th and this century, Third Age has evolved as a time in life when one is without familial or employment obligations but not yet dependent on others. It is a time when the young-old or Third Agers can reap the fruits of their labour usually after the onset of retirement, making this transition in the life course yet another defining characteristic of the young-old. Going back to life course perspective, Elder et al. (2003: 8) remind us what transitions such as retirement mean to individuals: “Transitions often involve changes in status or identity, both personally and socially, and thus open up opportunities for behavioral change.” Retirement is often (though not exclusi‐ vely) regarded as the onset of Third Age and the often-debated transition from 29 2.1 Towards a Definition of the Young-Old the Third Age to the Fourth Age is seen in the loss of independence due to factors such as illness. Laslett (1987) points out that retirement is an important, but not a necessary, condition for the Third Age, since some people can find personal fulfilment early in their lives meaning that, for some, Third Age can even occur much earlier in life, when others are still experiencing Second Age. This is where the definitions of Third Agers and young-old diverge, making the term Third Age less functional overall (Staehelin 2005: 167). Based on this and the discussion in the previous section, the cohort effects and the retirement factor constitute important characteristics of what I consider young-old. In this study, these are not the only characteristics but they are nevertheless more determining than chronological age. How do the young-old - similar to Laslett’s Third Agers - make use of their time of personal fulfilment and their maintained independence in retirement? As mentioned above, most of the young-old grew up expecting a three-box model life course. Saving and preparing for their Third Age, they also formed expectations of what it would be like once they are free from other obligations and hopefully still independent enough to do more travelling, gardening and - in the context of this study - engaging in lifelong learning (Hardy 2011). In other words, with these expectations of retirement, the young-old have formed pos‐ sible selves of their future retired self (see next chapter). This trend among the images of previous and current young-old cohorts has been picked up on an overall societal level and redirected to an image of optimal / successful / produc‐ tive / positive ageing: defying age-related diseases and maintaining an active lifestyle in later life (Daatland 2005: 375; Withnall 2012: 652-657; Rowe & Kahn 1997: 433; Phillips et al. 2010: 209-213; Schmidt-Hertha 2014; Schaie 2015). Due to the expanding life expectancy in relative good health and independence after retirement, and with it, the rise of the Third Ager and Young-Old concepts, expectations toward this phase of life and age group changed within industri‐ alized societies. Making the most of one’s retirement, maintaining quality of life for as long as possible, and ‘ageing successfully’ (Rowe & Kahn 1997; Schaie 2015) have become key issues for the young-old. Nevertheless, the goal to age successfully has also put more pressure on the individual. However, making lifelong learning obligatory (e.g. by employers or health insurances) could pos‐ sibly compromise an individual’s motivation to pursue lifelong learning (Pon‐ gratz 2003; Gronemeyer & Buff 1992; Whitnall 2012; Kade 2009). 30 2. Who are the Young-Old? 2.1.4 Psychological Dimension The psychological dimension of ageing looks at how our personality, our mental functions (e.g. our memory or intelligence), and sensory and perceptual pro‐ cesses change as we get older (Phillips et al. 2010: 12 f.; Lehr 2007). In particular, how old age and ageing are perceived in society and by the young-old them‐ selves is interesting for this study. This will also be discussed with regard to language learning in greater detail and from a more temporal perspective (i.e. in terms of past, present, future/ ideal possible selves) and in the following chapter. Regarding the perception and self-perception of the young-old, we can see how the components of the term - ‘young’ and ‘old’ - come into play in this and other dimensions; biological ageing, or the aforementioned bio markers such as wrinkles, grey hair, or sensory impairments indicate the ageing process and mark someone as ‘old’ (Karl 2009: 23). The chronological dimension together with the sociological dimension also denote a traditionally constructed old age by setting a point in time when a person moves from working and thus contri‐ buting to the economy, to a phase in life when a person withdraws from the workforce and receives a pension. From the sociological discussion of age, we can see how the interplay of these dimensions and developments define when a society regards an individual as old. These rules or perceptions that are es‐ tablished on a societal level also trickle down to the level of the individual’s perception of his or her age and the ageing process, making it a more internal dimension of the construction of age (compared to the sociological dimension of ageing). The psychological dimension of ageing takes a closer look at how individuals perceive and cope with the ageing process (Mlinac et al. 2011; Brandstädter et al. 1993; Ammon & Maehr 2008; Kotter-Grühn & Hess 2012). Several studies have indicated, for example, that older study participants usually feel younger than their actual chronological age (Smith & Ryan 2015: 311; Kade 2007: 14). With regard to resilience and coping with the ageing process, Mlinac et al. (2011: 4 f.) report that older people tend to rate others of the same age more poorly in terms of negative ageing stereotypes while perceptions of themselves improved thereby indicating a coping strategy: “This mechanism can be seen as a protec‐ tive factor; older adults reject negative information about aging rather than al‐ lowing it to harm their self-perceptions.” (2011: 5) The psychological dimension is also related to the ability of a young-old individual to adapt to new circums‐ tances and environments as he or she ages. Thus, going back to the sociological age factor of retirement, psychological gerontologists have divided this crucial transition in the life course into three phases: retirement planning, retirement 31 2.1 Towards a Definition of the Young-Old decision-making, and retirement transition and adjustment (Wang & Shi 2015: 349). How the young-old progress through these three phases of retirement - from envisioning their life in retirement, planning, and exiting their working lives to settling into their post-retirement life, ideally close to what they previously en‐ visaged - can tell us a lot about their ability to cope psychologically with change in later years. This information also helps them to maintain or even improve their psychological wellbeing. Dorfman and Kolarik (2005), for instance, report that engaging in bridging employment, volunteer work, and leisure activities (which can also include lifelong learning activities such as language learning) contribute to an improvement of psychological wellbeing in the transition and adjustment phase of retirement. It is thus not surprising that more and more young-old people engage in language learning as a way of improving and main‐ taining their post-retirement quality of life - i.e. enhancing personal agency, control, and emancipation (Ammon & Maehr 2008; Moody 2010: 485; Walker 2010; Whitnall 2012: 652; Ramírez Gómez 2016: 136). For adult learner institu‐ tions, this transitioning phase in life constitutes a promising starting point for new educational programmes (Schmidt 2005; Costard 2006; Tippelt et al. 2009; Völkening 2006). 2.2 The Young-Old Language Learner It is even on the cards that by the end of the next century the share of the oversixties not of adults but of the whole population may exceed half, and the expectation of life at birth rise into the nineties or higher. Can we really contemplate a future where a third or a half of all these people are regarded as the fag-end of the population, inactive, uncontributing, in limbo? - Laslett 1994: 444 - As mentioned above, the current cohort of young-old, comprised of members of the Silent Generation and Baby Boomers, have higher educational levels than previous generations. Correlating with this, there is also an increased interest in lifelong learning and, in the context of this study, an increased interest in post-retirement language learning (Berndt 2003; Whitnall 2012). Although a uniform view of adulthood in second and foreign language learning research still prevails, a few publications in recent years show the emergence of a distinct focus on older language learners and the first visible traits of a foreign language geragogy (Berndt 2003; Andrew 2012; Ramírez Gómez 2016). 32 2. Who are the Young-Old? In foreign language research, the majority of studies related to age have al‐ most exclusively focused on language acquisition from childhood to old age and, connected to this, a discussion of a critical period for language learning (e.g. Singleton & Ryan 2004). This type of research has - in concert with the de‐ ficit-model of ageing - added to the age-bias in society and the classroom, which also negatively impacts the self-image of older learners and thus their language learning performance (Andrew 2012; Ramírez Gómez 2016). It is for reasons like these that a discussion of a critical period will not be considered here. Other second and foreign language research often ignores older learners as a distinct group and instead focuses on more vocational-related issues of particular adult groups (e.g. migrants or university students) between 25 and 64 (European Commission 2016; Jarvis 2009; Dörnyei 2010; Mercer 2011a; Murray 2005). To understand the young-old language learner better, it is worth more closely examining what makes them start or continue learning a foreign language after retirement. The social aspect of attending language classes at a physical insti‐ tution is probably - next to the aforementioned intentions to maintain or enh‐ ance psychological well-being, maintaining a stable self-concept, and maintai‐ ning quality of life (Walker 2010; Wang & Shi 2015; Schmidt-Hertha 2014: 62 ff.; Staudinger & Heidemeier 2009) - one of the most important motivational factors (Murray 2011; Schnurr & Theisen 2009; Lang et al. 2012). However, as Ramírez Gómez (2016: 140) points out, this should not make instructors “overestimate the importance of socialization” and ignore the older learners’ main reason for attending a language class and thus neglecting the quality of foreign language instruction (on the outdated conflict between learning and socializing, see also Kade 2009: 59). As we will see in the data analysis (chapters 5 and 6), social reasons can differ in nature; a person can experience socialising within a com‐ munity of practice (i.e. joining a group with shared goals and identity at a lang‐ uage learning institution) - especially if one’s children have grown up and left the home (empty nest syndrome), or, if a person’s social network has become much smaller after leaving the workforce. These are some of the reasons and ways the young-old may look for ways to enlarge or to maintain their social network in these life-changing phases. Finally, it should be noted that studying the young-old is not without its challenges, some of which will be brought up again in the methodology chapter. The aforementioned minor role of the young-old language learner in language teaching and learning research means that there are few to no existing studies and research instruments available. I will discuss the ways in which existing research instruments need to be adapted or new instruments need to be deve‐ loped in chapters 5 and 6, hopefully making a crucial contribution by adding 33 2.2 The Young-Old Language Learner older learner-appropriate research designs and instruments to the field of lang‐ uage teaching and learning research. Moreover, in existing gerontological re‐ search the majority of studies are longitudinal in nature - to better capture developments of different cohorts and generations. This, however, requires more resources than cross-sectional studies. Cross-sectional studies, such as this study, take a ‘snapshot’ of the current situation of older age groups, which is why outcomes may be limited to the cohorts under investigation. Nevertheless, these outcomes can be a valuable resource to enable both young-old learners and foreign language teachers to become aware of the complexity and const‐ ructive nature of age (and self-concept), as well as providing both parties with strategies to enhance the foreign language teaching and learning processes. 34 2. Who are the Young-Old? 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners In this chapter, I hope to provide the reader with a better understanding of what self-concept is and how its temporal facets play a role in how the young-old in my study conceptualize themselves as foreign language learners. After a dis‐ cussion of the use and misuse of self-terminology and a look at what concepts mean on a more general level, I will attempt to clarify the meaning of self-con‐ cept and the aspects of it relevant to my research project by firstly, considering its structure and formation and, secondly, contrasting it with some commonly investigated self-terms in SLA research. I would then like to further narrow the focus on self-concept’s temporal as‐ pects, or more specifically, on the idea of the psychological construct of possible selves (Markus & Nurius 1986). How do we perceive the past, present, and future aspects of our self ? And related to this, how do we form ideal selves and ought-to selves over time and in what ways do they impact our life course? In the area of second language learning, Dörnyei (2005, 2009) has developed the Second Language Motivational System to find answers to these and more questions regarding the language learner self-concept. This and several other models in current Second Language Acquisition (SLA) research will be discussed in greater detail at the end of this chapter. Out of these models, Dörnyei’s Second Language Motivational System in particular will serve as the foundation for understanding the young-old language learner’s ideal and ought-to L2 selves. 3.1 A Note on Self-Terminology Of certain fallacies to which the psychologist is exposed by reason of his peculiar point of view - that of being a reporter of subjective as well as of objective facts, we must presently speak. - James, 1890: 184 f. - Due to a long history of research in the area of self, a “potpourri” (Klein & Gangi 2010: 2) of terms related to this phenomenon has evolved over the centuries. Most of the self-x terms, such as self-esteem, self-worth, self-awareness or self-confidence, we - as laypersons - tend to use interchangeably in everyday 1 „Monika needs to do something for her Self, Klaus still hasn’t found his Self, Inge seems to have severe problems with the Self, and overall it can be said that women still often have a weaker Self than men […]“ (author’s translation) parlance. This may be because we are constantly and naturally preoccupied with our self from an early age and throughout our lives (Hattie 1992: 1 f.; Marsh & Hattie 1996b: 446 ff.; Mercer 2014: 160). Mummendey (2006: 17) nicely illustrates this with an example for the usage of the term Selbst (self) by ordinary people on a daily basis: Monika muss mal wieder etwas für ihr Selbst tun, Klaus hat immer noch nicht zu seinem Selbst gefunden, Inge scheint erhebliche Probleme mit dem Selbst zu haben, überhaupt haben Frauen oft immer noch ein schwächeres Selbst als Männer […]. 1 According to him, the use of Selbst in these situations actually refers to various psychological phenomena such as a general change of lifestyle, self-esteem, or self-confidence. Mummendey (2006: 16 f.) points out that keeping in mind that these people are not professional psychologists, the way they (mis-)use the term self is no problem. When it comes to researchers in the field, however, the usage of self-related terms is a different story and he cautions researchers to avoid such reification of psychological phenomena. Even though taking caution with the use of terminology as a researcher is justified, it may in turn erroneously lead to thinking that the two worlds of lay persons’ usage and researchers’ usage of psychological terminology are to remain separate entities which are never meant to impinge upon each other. But research does not exist in a vacuum and those two worlds do in fact collide at times. This happens, for example, when laypersons are participants in re‐ search studies. Researchers such as Mercer (2014) have taken this issue into account by adapting instructions for their studies to participants’ everyday or lay understanding of the self or self-concept: “They [the pilot study participants] were shown the video and asked to evaluate their feelings of confidence. This self-construct was chosen because of its easy comprehensibility for the partici‐ pants and relatively broad scope.” (ibid.: 166) She used the term ‘confidence’ when instructing participants in her study when, in fact, she was investiga‐ ting ‘dynamics of the self ’. In what ways I have adapted my research instruments (questionnaire and interview) to the self-understanding of my participants, and how this may or may not have an impact on the ethical dimensions of my re‐ search will be discussed in chapter four. Going back to the latter quote by Mummendey, he nevertheless does make a valid point when warning about substantialisiertes Denkgebilde or “reification” (2006: 14). As with the usage of the term Seele (soul), which has been used more 36 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners widely in research before the 20 th century, Mummendey (2006: 18) sees a similar development of reification of the self in psychology, which may trick us into thinking of it as an independently acting entity. He then lists numerous examples of formulations in scientific articles and books, which imply an active role of the self. Other researchers in the field, such as William James (1890: chapter XII.) and John Hattie (1992), also remind their readers of this fallacy. In his book Self-Concept (1992: 242), Hattie takes up Ryle’s (1949) phrase ‘ghost in the machine’ to exemplify how not to approach psychological terms - in this case, self-concept: The intention is not to reify self-concept. Self-concept is not a ‘ghost in the machine.’ There is no hidden homunculus that sits inside our body that drives our view of the world. The psychological term ‘self-concept’ is no more real than are the notions of locus of control, intelligence, gravity, or electricity. Self-concept exists only in its ma‐ nifestations. Even though it may not always be possible to refrain from lending the words ‘self ’ or ‘self-concept’ a sense of agency in my formulations, the issue of reification is important to bear in mind when trying to grasp the aforementioned complexity of self-terminology. It is exactly this complexity which requires me - just like other researchers at the beginning of their research - to provide clear-cut boundaries of the terms self-concept and self within the framework of my research. Just as a surgeon needs a sharp knife to perform accurately, a theoretical foundation with sharp edges is required for easy navigation through one’s research project. So, to start sharpening my terminological edges in this chapter, I will give an overview of some central characteristics of self-concept by, firstly, summarising suggestions regarding its structure taken from relevant research literature; and, secondly, contrasting it with some commonly (mis-)used terms in the areas of second language acquisition (SLA) research and psychology. 3.2 Self-Concept is… Confronting the vast amount of self-terminology, it comes as no surprise that definitions of self-concept are not easy to put down on paper in a concise way. Self-concept has been referred to as beliefs (Mercer 2011a; Canfield & Wells 1976; Pajares & Schunk 2002; Jopling 2002), knowledge ( James 1890; Neisser 1988), a theory (Turner & Onorato 1999; Neisser 1988), a perception (L’Ecuyer 1975; Mercer 2011a; Mills 2014; Pajares & Schunk 2002) and cognitive appraisals (Hattie 37 3.2 Self-Concept is… 1992) about our self. At first sight, all of the above terms are attempts to find synonyms for the word concept. But do they express precisely and equally well what a self-concept is? To find an answer to this, I would like to take up the approach by Ulric Neisser (1993) and start by taking a look at the word con‐ cept. 3.2.1 Back to Basics: Concepts What is a concept in the first place? A concept, broadly defined, is what we associate with and what we believe - true or not - about something. This ‘so‐ mething’ can be concrete like a bird or a dog (“Can it fly? Does it have four legs? Does it have feathers or fur? ”); it can also be rather abstract and complex like war or peace (Neisser 1997: 4). Our concepts of these things can overlap, but are never exactly the same, and the extent to which concepts differ or are similar between different individuals can depend on different cultures ( Jopling 2002: 45; Hattie 1992). The network of concepts to which a single concept refers can be considered a ‘theory’ (Neisser 1988: 53). All of the above factors regarding concepts also apply to our concepts of the people around us and, of course, the network of concepts (or theory) we have about ourselves, i.e. our self-concepts. The important thing here is that self-con‐ cepts are more about what we believe to be true of our selves, or as Neisser (1993: 3) put it: Concepts are one thing, reality is another. The same thing can be said of your concept of Ulric Neisser. More important, the same thing can also be said of my concept of Ulric Neisser. Self-concepts never do full justice to the self. This makes self-concepts something subjective, something idiosyncratic; in other words, we need to keep in mind that the self creates a reality for an indi‐ vidual - one reality out of many. A reality of one’s self also diverges from how others create it from the outside. We can only assume that it is the sum of these created realities of self, which are shaping what is reality as Neisser describes. Since we as researchers are nowhere near being able to measure this objective reality, it is the subjective view of self of the young-old language learner that is methodologically more feasible to capture here in this study - even if not in its totality. In the light of this, ‘beliefs’ and ‘perceptions’ are the most appropriate terms to define self-concept in the context of my research, since they best cap‐ ture its subjective and dynamic nature. In the following, I will also use self-concept in both its singular and plural form, which - as we will see - highlights the fact that we can investigate 38 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners 2 However, Northoff (2014: 157) reports that at least the empirical self (in this case, the self as a biological process in the body and brain) can neurobiologically be located in the e.g. prefrontal regions of the brain; the pre-reflective (phenomenal) self as well as the minimal self could be located in the subcortical regions and cortical midline struc‐ tures. 3 In consonance with this, Oyserman et al. (2011: 94) have taken up this idea of self-con‐ cept being ‚nested’ in the self in their definitions and differentiation of self, self-concept and identity. self-concepts in different domains of our everyday life, but that there is at the same time a relatively stable part - the sum of our self-concepts, our theory of self - that is needed to give us a sense of unity (Canfield & Wells 1976; Markus & Wurf 1987; Jopling 1997). Are there other things we can agree on concerning human self-concept? Taking a closer look at different ideas of its structure and formation, which have been proposed by research so far, might present us with more universal characteristics of self-concept. 3.2.2 Of Pyramids, Jellyfish, and Molecules: Situating the Self and Self-Concepts in our Minds If, in short, there is a community of computers living in my head, there had also better be somebody who is in charge; and, by God, it had better be me. - Jerry Fodor, 1998 - Even though we should not imagine self-concept or the self to be some homun‐ culus in our heads, i.e. we should not reify the self, the following questions remain: Where are self and the self-concept located in our brains (and beyond)? How is self-concept structured? Again, answers to these questions cannot result in an exact description of an entity in our heads. 2 The answers will remain on a conceptual level, just as a structural formula of water is merely a graphic re‐ presentation of its molecular structure. In his work Principles of Psychology (1890), James provided us with one of the earliest attempts to illustrate our self-concept structure. James thought of the self to be comprised of the I-self (the knower; the subjective self) and the Me-self (what is known). What he la‐ belled Me-self has also been referred to as the objective self or the empirical self, and later as the self-concept (Harter 1996: 1 f.). 3 James further divided the Me-self into three parts: the spiritual self, the social self, and the material self (1890: chapter X). His formulation has influenced researchers after him (e.g. Shavelson et al. 1976; Song & Hattie 1984; Marsh & Shavelson 1985): He considered these three constituents to be in a somewhat hierarchical order based on their im‐ 39 3.2 Self-Concept is… portance for the individual - with a spiritual self at the top, a social self in the middle, and a material self at the bottom (Harter 1996). This idea of a self-con‐ ceptual hierarchy in our minds has been taken up, refined, and tested in nume‐ rous studies (for a detailed review see, e.g.: Marsh & Hattie 1996a; Byrne 1996a) - with Shavelson et al.’s model (1976) probably being the most prominent and influential. The hierarchy of the Shavelson et al. model (figure 2.1) was ordered in the following way: a general self-concept was at the apex of the model, which was then further divided into non-academic and academic self-concepts. Academic self-concepts in particular have been at the centre of attention in follow-up research and will also be relevant for this research project. It is also notable that their model implied multidimensionality (academic and non-academic factors; the academic dimension further included subjects such as math self-concept or English/ verbal self-concept) of the self-concept, in addition to a hierarchical structure. Both hierarchical and multidimensional aspects of the self-concept have been validated in numerous follow-up studies (for a detailed review see Byrne 1996b). The academic self-concept strand in the original Shavelson et al. model iden‐ tified by Marsh & Shavelson (1985) has, however, been revised after testing the model had indicated a close to zero correlation between math and English self-concepts (q.v. the following section). Based on empirical findings, the new model now had two separate higher-order academic factors (Math/ Academic Self-Concept and Verbal/ Academic Self-Concept) rather than one global aca‐ demic self-concept (1985: 120; Marsh 1990; Xu et al. 2013). From her review of these different validation studies, Byrne draws the following conclusion: These findings underscore the domain specificity of academic self-concept and, thus, the need to use instrumentation designed to measure specific rather than global di‐ mensions of academic self-concept research that bears on perceptions of self relative to particular academic subject areas. (1996b: 292) Clearly, self-research has moved from investigating one global self-concept to viewing and measuring it in a more multidimensional and domain-specific way (Markus & Wurf 1987: 301; Marsh et al. 1988; Swann & Bosson 2010); a hierar‐ chical and multidimensional self-concept whose facets are becoming more dif‐ ferentiated and independent with age (Shavelson & Marsh 1985). But to what extent do we actually “run around with pyramids in our heads” (Hattie 1992: 242)? Instead of a two-dimensional or “monolithic” (Mercer 2011a: 70) entity in our minds, other researchers have argued for three-dimensional representations of 40 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners our self-concept (e.g. Bracken 1996). The introduction of the book A 100 Ways to Enhance Self-Concept in the Classroom (1976) serves as an early, intuitively accessible example of this. Here, the authors Canfield and Wells depict self-con‐ cept as a glob with dots in it, or, as they write, a “strong jellyfish” (p. 2). With the help of the dots’ positions, they explain the role of core and peripheral beliefs within our self-concept: Notice that some of the dots are more ‘internal’ than others. The more central a belief is to your Self, the more value - either negative or positive - you attach to it. For example, most teachers feel that to be smart is more important than to be a good swimmer. (Not everyone feels that way.) (1976: 2) They argue that the more central or internalized a dot (belief) is, the more dif‐ ficult it becomes to change this belief. Markus and Wurf (1987: 302) also make this differentiation between core and peripheral conceptions based on their im‐ portance with regard to their overall sense of self. They consider core concep‐ tions to be well elaborated and, thus, to have a higher impact on our behaviour than peripheral conceptions. This makes the core beliefs less susceptible to change than peripheral, i.e. less stable, self-beliefs. Another alternative model - similar to that by Canfield and Wells’ glob or jellyfish - has been developed by Mercer (2011a: 68). Her model focuses speci‐ fically on the self-concept domain of ‘language learning’. Based on the data of the participant “Joana”, Mercer created a molecular model or network model of Joana’s self-concepts (e.g. Physical Self-Concept, Academic Self-Concept, Ge‐ neral Languages Self-Concept, English as a Foreign Language Self-Concept). The advantages of this approach to modelling self-concept are that multiple con‐ nections between different facets or domains of self-concept can be made. The strength of connections as well as the size (importance) of individual facets can be highlighted in the ‘molecule’ and these give a more realistic impression of self-concept complexity (ibid.: 67). In this model Joana’s English (EFL) and Italian (IFL) as a Foreign Language self-concepts are not only quite well elaborated and complex (= size of spheres), but also have a strong connection (= size of con‐ necting line between spheres) based on the collected data. Nevertheless, Mercer highlights that this representation of Joana’s self-con‐ cept is only valid for this particular part of her self-concept, in the context of foreign language learning and at the time of data collection. This in turn means that this type of model is able to capture the dynamic and contextual nature of self-concepts (2011a: 68 f.). This leads to the next question about self-concept: How do we develop or form self-concepts? 41 3.2 Self-Concept is… 4 If and in what way our self shapes and possibly distorts reality will be discussed in section 3.3.1. 3.2.3 Self-Concept Formation People often say that this or that person has not yet found himself. But the self is not something one finds, it is something one creates. - T. Szasz 1973: 49 - As we grow up, we are accumulating more and more information about our self. Every new piece of information is scanned with regard to our already established perception of self. It is then either confirmed and internalized, or disconfirmed and discarded (Canfield & Wells, 1976: 2). Markus and Wurf write: “When sti‐ muli, experiences, or events cross the threshold of self-concept such that they achieve self-relevant meaning, they become special” (1987: 328). As we grow up and grow older we learn to filter incoming information (“stimuli, experiences, or events”) in a personalized fashion. 4 As mentioned in the previous section, self-concepts are stable and dynamic at the same time and thus peripheral concepts are more susceptible to change than core self-concepts (Markus & Wurf, 1987). Sometimes they can even be‐ come part of our core self-concept and change it. On the one hand, as we are learning more about ourselves through experience, this shapes our current self-concept accordingly. On the other hand, Neisser (1997: 4) writes “the op‐ posite direction of causation is also important: If I behave in certain ways in order to live up to an idealized self-concept, for example, my real self is being partly shaped by my conceptual self.” In other words, self-concept is not the passive receiver of our experiences but also has the power to make changes to the core perceptions if deemed necessary. What do I mean by writing ‘if deemed necessary’? I wrote at the beginning of this section that we scan, confirm, or disconfirm pieces of incoming information. But I have not discussed where this information comes from and how we decide whether it contains, as Markus and Wurf (1987: 328) wrote, “self-relevant meaning” or not. Rubio (2014: 43) lists some of the most important factors which help us to determine this: “The process of per‐ ception for the self-concept, as happens when looking at a picture, can vary depending on the personal maturation of the self, age, personality, situational conditions, metacognitive strategies, and many other individual contextual fac‐ tors.” What Rubio does not explicitly mention here are if and in what way other factors such as ‘gender’ and ‘significant others’ (in the case of young-old lear‐ ners, significant others could be other learners, the teacher, but also partners, 42 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners children, grandchildren, and friends) can influence our self-concept. Some of the factors listed by Rubio and these two additional factors, which will be im‐ portant for my research context, need further discussion. In Rubio’s quote we can detect the premise of a dynamic or changing self-con‐ cept, which is dependent on the process of perception in different situations and contexts. Related to this, Shavelson et al. (1976: 412 ff.) postulated that within the framework of their hierarchical model, self-concept facets become more dy‐ namic (or less stable) when moving downwards from the apex, i.e. when moving from “General Self-Concept” down to “Evaluation of Behavior in Specific Situ‐ ations” (in the case of EFL, this specific situation could be the acting out of a role-play together with another learner) in their model. According to Markus and Wurf (1987: 328), self-concept is considered to be dynamic in that it is “an active, interpretive structure that is continually involved in the regulation of on-going behaviour” - which they have depicted in their model of the dynamic self-concept. They support the view that there is not a single, unitary self-con‐ cept but that it rather consists of active and non-active parts. Markus and Wurf (1987) refer to the active parts of our self-concept as the working self-concept (see also: Markus & Nurius 1986) and describe it as “conti‐ nually active, shifting array of accessible self-knowledge” (p. 306). The working self-concept comprises self-representations, which have been or can be activated through internal processes (e.g. motivational state) and/ or through current so‐ cial situations, which is why they also refer to it as “self-concept of the moment” (ibid.). In their model, self-concept (core) is a collection of self-representations, which are - in contrast to their subset in the working self-concept - not acces‐ sible in a particular situation (ibid.: 314). It also illustrates how our working self-concepts affect - and are affected by - intrapersonal behaviour and inter‐ personal behaviour. Markus and Wurf (1987: 306) conclude: “The working self-concept thus consists of the core self-conceptions embedded in a context of more tentative self-conceptions that are tied to the prevailing circumstances.” Based on this, their model can be linked to the symbolic interactionsts’ view that self-concepts are greatly affected by social conditions. Additionally to that, it also accounts for the, at first, paradoxical sounding claim that our self-concepts are stable and malleable at the same time (ibid.). Age and our ageing process also affect the development of our self-concepts. We start to form our self-concepts quite early in our lives, i.e. towards the first year (Neisser 1993: 5; see also: Tomasello 1993). And, as mentioned before, self-concepts become increasingly differentiated and multidimensional, as we grow older (Shavelson et al. 1976; Byrne & Shavelson 1996; Leung et al. 2013). 43 3.2 Self-Concept is… The hierarchical model reflects this by the number of different domains in aca‐ demic and non-academic self-concepts. It should be noted, though, that this process does not mean we only diffe‐ rentiate our self into more unrelated, distinct self-concepts, but that we simul‐ taneously integrate similar or related self-concepts. In the words of Marsh and Ayotte (2003): “with increasing age and cognitive development, there are coun‐ terbalancing processes of integration and differentiation” (p. 702). The notion of a more differentiated self-concept with increasing age, however, has mainly been tested and validated for children, pre-adolescents, and adolescents (cf. Byrne 1996a). As we will see in chapter 5, the lack of appropriate research in‐ struments (e.g. questionnaire items that consider the age-specific situation of individuals post-retirement) for older participants is still an issue. Our cognitive development (maturation) has an impact on how we deal with incoming infor‐ mation to develop our self-concept. The types of information we receive from outside - i.e. the question of what we receive - also shape our self-concept. This external information can come from our own actions but also from other people around us. Thus, we need to ask next: What roles do significant others around us play in shaping our self-concepts? The two symbolic interactionists Cooley (1902) and Mead (1922) highlighted the idea that the formation of self is a social process. Cooley’s looking glass self, summarized in his couplet “Each to each a looking glass, reflects the other that doth pass” (1922: 184), illustrates that we construct our selves based on our perceptions of how people around us perceive us (see also: Oyserman & Markus 1998; Swann & Bosson 2010). We internalize the appraisals of significant others. Bouchey and Harter (2005), for instance, based their study on this premise. They revealed that reflected appraisals of parents and teachers (“socializers”) influ‐ ence students’ academic self-perceptions during early adolescence. Interes‐ tingly, perceived reflected appraisals of adults even had a stronger impact than those of their peers (ibid.: 682). Since the young-old are not as dependent on “socializers” as adolescents, does this mean that they are more independent in forming their self-concepts, i.e. the looking glass is no longer needed? Or, is it that the perspectives are reversed? Are many of the significant others shaping the young-old’s self-concept now part of younger generations: children, grandchildren, and - quite often - their English teachers? To the best of my knowledge, research is still needed to in‐ vestigate if there is still an impact of this reverse situation on self-concept. Another study by Harter and Whitesell (2003) points to the case of ‘major life transitions’ and their influence on self-concept formation. In their study, they noticed that even when moving on from high school to college, approval re‐ 44 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners 5 Self-Worth is often used interchangeably with self-esteem and refers to one’s perceived value as a person (Rubio 2014: 42). Self-concept is regarded by some researchers as perceptions of self plus judgments of one’s self-worth (Mills 2014: 10; Pajares & Schunk 2002: 21). ceived from significant others had an impact on participants’ global self-worth. 5 They posited: With regard to the applicability of Cooley’s formulation, one may question whether external approval should have such an impact if, over the course of development, one presumably internalizes the perceived approval of significant others such that one is no longer dependent on this type of social feedback. (p. 1041) Major life transitions (in their context: educational transitions) seem to explain why individuals turn to Cooley’s looking glass once more to reassess their self-concepts: The point is that looking-glass self-processes are appropriately reactivated in new academic and social settings, although such activation is unlikely to persist as students develop a new understanding of the self in the face of the opinions of those in the new campus culture. (ibid.) With regard to the young-old then, we need to ask if and to what extent retire‐ ment, as a major life transition in the life course (see chapter 2), has a similar influence on self-concept formation. Leaving the workforce changes our social environment when we, for example, no longer see our former colleagues that often (if at all). Retirement leaves us with more free time that many retirees feel required to structure with daily and weekly routines. Among these are often new activities that have been either put off or neglected for a long time such as pilates, a computer class, starting to learn Italian, or refreshing English language skills. The new challenges and new people (classmates and teachers) the young-old meet shortly before, during, and after the transition from work to retirement have the potential to re-shape their self-concept (see also Mercer 2011a: 152 f.). The role of others in forming our self-concepts is a subject that was also taken up by Marsh (1986) when revising the model of Shavelson et al. (1976). As noted in section 3.2.2, when testing the model in various studies, researchers were puzzled by the close to zero correlation between math and English self-concepts (Marsh 1986), even though they predicted all academic self-concepts on that level correlate to at least some extent. Marsh explained this with the Internal/ External (I/ E) frame of reference model theory. According to his theory, students 45 3.2 Self-Concept is… formed their verbal and math self-concepts based on internal and external frames of reference (ibid.: 133): Students compare their perceived math performance with their perceived verbal performances and other academic areas (internal). They compare their perceived math and verbal performance to the perceived abilities of their peers in one particular subject (external). To illustrate this, Marsh gives the example of a student who correctly per‐ ceives that both his or her verbal and math performances are below average compared to peers’ performances in these areas (i.e. external comparison), but who also perceives his or her math abilities to be better than his or her verbal ones (i.e. internal comparison). Then, according to Marsh, “[d]epending on how these two components are weighted, this student may have an average or even an above-average self-concept in mathematics despite his or her poor math skills.” (ibid.: 133) Linked to this, researchers also reported differences in self-concept based on gender, with female learners favouring verbal and male learners favouring math self-concepts (Marsh 1986; Hattie 1992: 176-180; Mercer 2011a: 80-83). It is arguable, however, if and to what extent these gender-biased beliefs regarding verbal and math domains are existent in the cohort of young-old in this study. Mercer (2011a) has used the I/ E frame of reference in her study on language learner self-concept in order to illustrate the self-concept formation of univer‐ sity students. Within her study, the internal frame of reference focuses on com‐ parisons across other domains (e.g. other foreign languages learned) and learner beliefs regarding learning a foreign language (ibid.: 125). The external frame of reference refers, for example, to social comparisons, feedback from significant others, (perceived) experiences of success and failure, and critical incidents (ibid.: 161). Both internal and external frames of reference - though not defined as broadly as in Mercer’s study - have also proved to be significant in the ana‐ lysis of this study’s interviews (see chapter 6). The findings of Mercer’s study have also indicated a change of L2 self-concept over time, experience, and lear‐ ning context (i.e. the transition from school to university). Thus, it is hoped that the findings of my study of young-old language learners which uses a mixed methodology might expand these insights to later stages in life (preand post-retiring) and other learning contexts (vhs centres). What the overview shows once more is that the majority of studies on the learner self - here more specifically related to the I/ E frame of reference theory - are focused on younger age groups. It would therefore be of great interest to expand the idea of internal and external comparisons to the learning context of young-old language learners. Do the young-old compare their language learning 46 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners 6 Mummendey (2011: chapter 3), for instance, discusses 25 different German and English self-terms and Leary and Tangney (2012: 10) list 66 different self-related terms. performance to their previous learning experiences (internal) and to the per‐ formance of their young-old classmates (external)? Thus, in particular, the in‐ terview conducted for this study contained questions about the language lear‐ ning group and other learners in the group (see chapter 6). 3.2.4 Definition by Demarcation: What Self-Concept is Not In the following, I would like to further sharpen the terminological edges of self-concept by contrasting it with other self-related terms. Since there are nu‐ merous self-terms 6 , I will focus on those most commonly discussed in today’s SLA-research, namely self-esteem, self-efficacy, and self-confidence. In what ways do these terms differ from, and overlap with, self-concept? 3.2.4.1 Self-Esteem William James (1890: 310) gave a very early, if not the earliest, definition of self-esteem in the mode of a formula: pretensions/ success = self-esteem Based on this formula, we form our self-esteem in a particular area of our life or domain (e.g. learning a language) based on how highly we value it or aim at succeeding in it. If a student has strong aspirations (i.e. pretensions) to, for example, be good at speaking English, then, because he or she thinks that this is important, success or failure in this endeavour affects his or her self-esteem more strongly in this particular domain. While self-concept is what we believe about our self, self-esteem involves an evaluation - positive or negative - of what makes up our theory of our self. In other words, self-esteem is a subjective evaluation of one’s worth (Orth & Robins 2014: 381). Rubio (2014: 42) compares the difference between self-concept and self-esteem using the metaphor of a picture: “Self-esteem would correspond to the resulting evaluation of the picture; while the picture would be the self-con‐ cept.” The keyword when it comes to differentiating the two terms here again is “evaluation.” 3.2.4.2 Self-Efficacy The slogan “Yes, we can! “ used in the US presidential elections in 2008 nicely captures what self-efficacy is all about. The way we evaluate our capabilities to perform a task depends on our self-efficacy regarding that particular 47 3.2 Self-Concept is… task: “Self-efficacy differs from self-concept in that it is concerned less with the skills and abilities one thinks one has but more with what one can do with whatever skills one possesses.” (Skaalvik & Skaalvik 2008: 224). Linked to this, self-efficacy can also be considered to be more cognitive than self-concept in that it contains more concrete or explicit judgments of one’s own perceived abilities (Mercer 2011a: 15; Leary & Tangney 2012: 10). Self-efficacy can guide and predict people’s behaviour, motivation, and deci‐ sion-making. Thus, it can influence the way we think about our abilities, e.g. whether a language learner approaches the experiences of language learning with an optimistic or pessimistic mindset (Bandura 2001, 2012). But prior to that self-efficacy already comes into effect: when making the decision to attend a language class, a person is convinced that she has what it takes to learn a lang‐ uage. Bandura (2001: 10) writes: “It is partly on the basis of efficacy beliefs that people choose what challenges to undertake, how much effort to expend in the endeavour, how long to persevere in the face of obstacles and failures, and whe‐ ther failures are motivating or demoralizing.” Whether at the moment a person takes on language learning or in the midst of the common ups and downs of the language learning experience, self-efficacy beliefs continuously act as guides. Several researchers see the difference between self-concept and self-efficacy beliefs in terms of specificity. Self-efficacy is measured at a more contextor task-specific level than self-concept (Mills 2014: 11; Pajares & Schunk 1994: 194). Mercer (2011a: 15) orders the three concepts of self-esteem, self-concept, and self-efficacy on a scale from global (self-esteem) to tightly context-specific (self-efficacy), with self-concept being in-between. Bandura (2012: 17), however, points out that in many studies the understanding of self-efficacy is often too narrow, being merely on a task-specific level. While he does not doubt the va‐ lidity or accuracy of studies measuring self-efficacy, he nevertheless reminds us that self-efficacy should be measured in terms of ‘activity domains’ rather than of reducing it simply to the level of tasks. He illustrates this as follows: [I]n the standard self-efficacy assessment, an activity domain is measured by a set of specific items representing gradations of challenge, but the self-efficacy score is for the totality of items encompassing the domain of activities, not for a specific item within it. […] It is time to retire the misleading claim that self-efficacy theory is in‐ herently wedded to ‘narrow measures’. (2012: 17) What he calls “activity domains” can encompass different degrees of complexity. It is, for example, possible to measure efficacy beliefs regarding learning a new language (broad scope efficacy) or to measure efficacy beliefs regarding making a booking over the phone in another language (limited scope efficacy). This 48 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners means that it is more difficult than anticipated at first sight to differentiate self-concept and self-efficacy measures in terms of their specificity (Mercer 2011a: 15). Concluding from this, the primary remaining difference between the two concepts lies in the fact that self-efficacy is more closely bound to “can-do”-appraisals than self-concept. Measuring via a can-do appraisal in this study by asking if a young-old participant can, for instance, envision himor herself communicating successfully with a native speaker in the near future (see chapter 6) tells us primarily something about his or her self-efficacy regar‐ ding that task. Nevertheless, we can draw conclusions - together with the results of similar questions - regarding his or her self-concept in the domain of using (or speaking) English in the future. 3.2.4.3 Self-Confidence Self-confidence is another self-term often used interchangeably with other self-terms in everyday life. As previously mentioned, Mercer made use of its commonality when investigating the L2 self. When interviewing her study par‐ ticipants, she used the more lay-friendly term ‘self-confidence’ when, in fact, she was actually researching the more complex concept of ‘L2 self ’ (2014: 166 & 173). What exactly constitutes self-confidence - or, in this case, more speci‐ fically L2 self-confidence (L2C) - and how it is developed, shows in what way it differs from self-concept. Clément has done extensive research on L2 self-confidence. Together with Sampasivam (2014: 25), he refers to L2C as a “higher-order variable encompas‐ sing both a lack of anxiety and positive self-ratings of L2 proficiency.” On a more general level, Dörnyei (2005: 73) defines self-confidence as a “belief that a person has the ability to produce results, accomplish goals, or perform tasks compe‐ tently.” At first glance, this definition of self-confidence is reminiscent of self-ef‐ ficacy. According to Mercer (2011a: 17), however, L2 self-confidence is less do‐ main-specific than self-efficacy. Moreover, Dörnyei (2005: 73 f.) adds that self-confidence is less cognitive than self-efficacy and primarily socially const‐ ructed. Indeed, when it comes to the development of self-confidence, Sampasivam and Clément (2014: 25) report that self-confidence develops through “contact experiences.” Based on this, context of L2 acquisition plays an important role for the developing L2 self-confidence. Sampasivam and Clément (2014: 26) conc‐ lude that it is worthwhile to organize L2C literature according to learning out‐ side and inside the classroom. Previous research suggested that “it is considered to have a less influential role in communities with infrequent intergroup contact 49 3.2 Self-Concept is… 7 Volkshochschule (lit. transl.: folk high-school - from here referred to as vhs centres) is a German adult education organization offering a wide range of continuing education classes (languages, crafts and art, sports, politics) to the public (regardless of age or degree) at centres all over Germany (see chapter 4). compared with multilingual communities. Nevertheless, the quality and quan‐ tity can have different effects.” (p. 25) In the case of the young-old learning at Volkshochschulen (hereafter vhs centres) 7 in Germany, which can be considered a foreign language learning (FLL) context, contact with native L2 speakers is rare. Does this mean that learners’ self-confidence does not develop as well in a FLL context as it would develop in a SLA context due to a lack of “contact experiences” with L2 community mem‐ bers? Clément et al. (1994: 42) write that language self-confidence and language anxiety can be linked to L2 classroom experiences (e.g. evaluations by the tea‐ cher or other learners). Moreover, Sampasivam and Clément (2014: 27 f.) report studies which “suggest that in the FLA context, foreign media use and perceived importance of contact were better predictors of motivated learning behaviours (items concerning the different aspects of students’ learning behaviour) than more direct forms of contact.” The young-old learners who are learning a lang‐ uage at vhs centres in Germany can form their self-confidence independently of direct contacts with the L2 community. Classroom experience (e.g. external frame of reference; see chapters 7 & 8) and indirect contact through such things as media, as well as the degree of importance they place on these aspects of language learning, shape their feelings of anxiety and self-confidence. Mercer (2011a: 17) also contrasts self-concept and self-confidence respectively in regard to the focus of their affective dimension: It thus seems that whilst L2 linguistic self-confidence reflects some aspects of self-con‐ cept, such as domain-specific beliefs of competence, the affective dimension differs as it focuses more on feelings of anxiety, whereas the affective aspect of self-concept is more typically concerned with evaluative feelings associated with the self-beliefs in the domain. Summing up this short discussion of L2 self-confidence, the key differences in regard to self-concept lie in, firstly, its development through direct and indirect contact experiences (inas well as outside the L2 classroom) and, secondly, its more specific link to anxiety. 50 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners 3.2.5 Summary In the above, I have discussed what we need to consider when using self and self-concept terminology. I have introduced several perspectives on self-concept structure and its formation, and finally have contrasted self-concept with some other self-terms, which are commonly investigated in SLA research but often confused, in order to create a clearer picture of what self-concept is, or is not in the context of this study. In the following, I would like to sum up the discussion in this section and come to my understanding of self-concept. • The conceptions of our self are subjective. From the various attempts to describe or paraphrase self-concept, the two key terms perceptions or beliefs highlight particularly well how our self-concepts create a subjec‐ tive reality of our selves. The way a young-old conceptualizes his or her language learner self diverges and overlaps with how others - teacher and classmates - may view it, creating a dynamic between different sub‐ jective realities. Whether we are talking about one general self-concept or about a multitude of self-concepts: we are dealing with different, do‐ main-specific self-concepts, such as foreign language learning, but also at the same time with a unifying, core sense of self. The core and peri‐ pheral facets of self-concept also come into play when taking a look at its structure and formation. • Self-concept has a hierarchical, multidimensional structure, which becomes increasingly differentiated with age. Inspired by William James’ earliest works on psychology from the 19 th century, the hierarchical structure model of self-concept has dominated research until today. Of particular interest for my research project is the academic strand of Shavelson et al.’s version of this model (1976), which not only depicts the hierarchical order of self-concept facets (with the general self-concept at the apex resembling a core or relatively stable part), but also its multidimensiona‐ lity (e.g. self-concept across different academic domains). Within the verbal academic dimension of self-concept (including foreign language learning), for example, Mercer (2011a) approached depicting different fo‐ reign language self-concepts in a three dimensional model. Unlike the hierarchical model, this sphere model is able to capture the contextual and dynamic nature of and, more importantly, between different foreign language self-concepts. • Self-concept has a relatively stable core and more dynamic peripheral parts. In what ways is self-concept dynamic and contextual? Markus and Wurf ’s model (1987) takes up the symbolic interactionist’s notion that forming 51 3.2 Self-Concept is… one’s self-concept is a social process. It also takes into account that how we ourselves perceive our actions and appraisals shapes our self-concept. Both interand intra-personal factors affect how our working self-con‐ cept either internalizes or discards incoming, potentially self-relevant, information. In the case of the young-old language learners, how they perceive their own language learning process and progress (intra-per‐ sonal factors), as well as how they ‘filter’ incoming self-relevant infor‐ mation from the teacher, other language learners in their class, and people they communicate with in the target language outside the classroom (inter-personal factors), form the basis of their language learning self-con‐ cept. Markus and Wurf ’s model shows how self-concept can be stable and dynamic at the same time. Finally, what we need to keep in mind is that self-concept formation is a bi-directional process between our real and idealized self. A young-old language learner’s self-concept grows out of the tension of his or her perceived status quo and his or her envisioned ideal language learner self, guiding his or her further language learning process. • Self-concept differs from other self-related constructs in the degree of cog‐ nition, evaluation, and specificity involved in its development and measu‐ rement. It is not always easy to keep clear-cut boundaries between the commonly addressed concepts in SLA research such as self-concept, self-esteem, self-efficacy, and self-confidence when trying to measure or to investigate them. Contrasting them makes boundaries and overlaps visible. On a scale ranging from highly evaluative to highly cognitive, self-esteem is more on the evaluative side than self-concept, which is more cognitive in nature. Self-efficacy is even further down on the cog‐ nitive side of the scale compared to self-concept. What remains somewhat open to debate and difficult to differentiate when measuring is the degree of task-specificity of self-concept vs. self-efficacy, with the latter being considered the more task-specific one. Self-confidence, another often in‐ vestigated psychological concept in SLA-research, differs from self-con‐ cept in that it is more strongly connected to anxiety and has its roots mainly in contact experiences with the L2 (direct and indirect). Thus, in many cases a single item or a group of items in the questionnaire for this study as well as an interview question can indeed point to several self-as‐ pects of a young-old language learner. For instance, asking a young-old learner to rate her English skills on a scale from 1 to 10 together with an elaboration of the rating, can help to draw conclusions not only about her self-concept but also about, for example, her self-esteem. 52 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners Finally, to illustrate all of the central characteristics of self-concept that I have discussed above in a more condensed way, I would like to draw on Hattie’s (1992) facet analysis of self-concept (see figure 3.1). While he uses the term cognitive appraisals to describe self-concepts (alongside expectations, descriptions, and prescriptions), I have used the terms beliefs and perceptions. The integration across various dimensions and the ways in which this happens (self-verification etc.) corresponds with the multidimensional, hierarchical structure (self-con‐ cept structure) and the internalization process (self-concept formation) dis‐ cussed in sections 3.2.2 and 3.2.3. The last part of his facet analysis refers to the stability vs. dynamics of self-concept (the core and peripheral parts of self-con‐ cept discussed in section 3.2.3) as well as how we use intraand inter-personal information (dis-/ confirmation) to (re-)shape our self-concept. self-concepts Our conceptions of self are cognitive appraisals, expectations descriptions expressed in terms of prescriptions, integrated across various dimensions that we attribute to ourselves. self-verification self-consistency self-complexity The integration is primarily via self-enhancement. consistent type These attributes may be inconsistent depending on the amount of confirmation others disconfirmation our appraisals receive from ourselves. Figure 3.1 Facet analysis of self-concept by Hattie (1992: 37 - bold by me) Arriving at an overall understanding of what constitutes self-concept and how it is formed, I would like to focus on a more specific area of self-concept in the next section: its temporal facets. As mentioned above, the dynamic between actual and ideal self is a driving force in a young-old learner’s language learning 53 3.2 Self-Concept is… 8 Taylor (2014), in the same publication, makes a similar differentiation of synchronic and diachronic views of self, with the latter being an equivalent of temporal self. process. What roles do the past, present, and future selves have on young-old language learners? What can we learn from investigating their possible, ideal, and ought-to selves in order to find out how to facilitate their language learning process? Second language acquisition research has already focused on the tem‐ poral aspects of self-concept and possible selves, resulting in different models that help to understand the language learner self from a temporal and motiva‐ tional perspective (see section 3.4). The next section on the temporal self pre‐ sents and discusses the foundation for SLA models of language learning and the self. 3.3 The Temporal Self What’s past is prologue. - Shakespeare, The Tempest, act II, scene I, lines 253-54 - Mercer and Williams (2014), in their concluding remarks, divide research on the self in the SLA classroom into two broad categories: the contextual self and the temporal self. 8 Based on this distinction I would like to focus on the latter - the temporal - aspect of self-concept in this section. This is not to say that context or the contextual self do not play a role in my study, as we will see in the next chapter. I focus on the temporal aspect for two main reasons: Firstly, as we will see in this and the following section, the temporal perspec‐ tive on the self in psychology and SLA research lays the foundation for central theories and models used to explore the self. In SLA research, for example, Dörnyei ’s (2005) L2 Motivational Self System (L2MSS) draws on future-oriented self-aspects such as ideal and ought-to selves of L2 language learners (see section 3.4.1). Secondly, for no other younger age group does time play such a central role as for the young-old (Carstensen 2006; Koutoukidis et al. 2013: 234; Russell 2011). They can look back onto a longer past than younger age groups, with various experiences to feed their current perception of self and construct future-oriented selves. Simultaneously, this particular phase of life is connected to major life changes, such as retirement, children having moved out from home (‘empty-nest syndrome’), or higher death rates among friends of their age, which trigger an increased awareness of time and the limits of life. 54 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners 9 Connected to this debate is the issue of measuring or investigating the present or actual self, but also the dynamics and stability of self and the working self, as discussed in section 3.2.3. A first basic distinction of the temporal self can be past, present, and future self. Pinpointing the present, where it begins and where it ends (and inextricably linked to this are the beginnings and ends of past and future), is debatable. Is the present, and with it the present self, just that millisecond moment slicing the past and the future? 9 Keeping this in mind, I am going to introduce and discuss three different constructs, which subsume influences of our past, pre‐ sent, and future selves: possible selves (Markus & Nurius 1986), self-discrepancy (Higgins 1987), and self-determination (Deci & Ryan 1985). 3.3.1 Possible Selves Markus and Nurius (1986) define possible selves as self-representations of what we might become, what we would like to become (ideal selves) or what we are afraid of becoming (feared selves). The authors link possible selves to the dy‐ namic nature of self-concept, i.e. motivation, distortion, as well as momentary and enduring change, and thus see it as “a conceptual link between cognition and motivation” (ibid.: 954), or, in the words of Dörnyei (2009: 11), “how the self regulates behaviour by setting goals and expectations”. This means that possible selves function as a major impetus to behave in a certain way in order to, on the one hand, move closer towards a particular ideal self, e.g. “the successful self, the creative self, the rich self, the thin self, or the loved and admired self ” (Markus & Nurius 1986: 954). In the case of a young-old language learner an ideal L2 self could be conversing fluently with other tourists and L2 speakers at the hotel bar, or feeling more at ease when partaking in speaking activities in class. On the other hand, possible selves make us move or stay away from a particular feared self: “the alone self, the depressed self, the incompetent self, the alcoholic self, the unemployed self, or the bag lady self ” (ibid. 954). Again, regarding our young-old language learners, this feared self could be the L2 self who is not understood at the hotel reception, or who fails to formulate requests in the L2 when booking over the phone. Based on this linking of cognition and motivation illustrated above, possible selves are also often referred to in the literature as guides. Ryan and Irie, for instance, describe them as self-representations which “guide individuals to‐ wards becoming the person they would like to be, a hoped-for self, or away from outcomes they seek to avoid, a feared self” (2014: 112). Similarly, Dörnyei and Ushioda speak of future self-guides that reflect a “dynamic, forward-pointing 55 3.3 The Temporal Self 10 Markus and Nurius (1986: 954) use the term “salient others“ when writing about how possible selves are socially derived. conception that can explain how someone is moved from the present toward the future” (2011: 80). The latter view, however, has led some researchers to put a strong emphasis only on the future when discussing possible selves. Different from this view, then, it is necessary to remember that possible selves are not only linked to the future: “Possible Selves derive from representations of the self in the past and they include representations of the self in the future. They are different and separable from the current or now selves, yet are inti‐ mately connected to them” (Markus & Nurius 1986: 954). Past and future possible selves are mirrored in current representations of self (Ryan & Irie 2014: 111). Past selves may become relevant in the future again: The past self of a student who never did her homework or did not do well in English tests at school may influence the perception of her current self as well as the formation of a possible future self which this young-old learner may be afraid of again becoming when attending English classes decades later. But the future possible self of a young-old language learner could also be one she hopes to become again, if, for example, she used to be an A-student in English and later experienced successful communication using the L2 with others in her job or on holidays. Consideration for the fact that young-old learners in our classes each have long and quite varied pasts gives us an idea of how both multifarious and personalized their possible selves must be. The view that possible selves are unique to an individual, i.e. personalized (Markus & Nurius 1986: 954), is linked to the discussion of self-concept forma‐ tion (see section 3.2.3). One of the reasons lies - once more, similarly to self-con‐ cept (looking-glass self) - in a person’s socio-cultural history and context as well as the influence of significant others: 10 “Possible selves thus have the potential to reveal the inventive and constructive nature of the self but they also reflect the extent to which the self is socially determined and constrained” (ibid. 954; see also Oyserman 2008). If we apply this to a specific age group such as the young-old, it becomes clear that cohort effects (socio-cultural, historical, poli‐ tical and other impacts on a cohort) play a role in the formation of their possible selves. On a more general level then, possible selves exist for a whole age group or cohort, down to more personalized influences on the individual’s possible selves. But future-oriented possible selves develop on the basis of an individual’s history and are therefore to a great extent also personalized. To illustrate this cohort effect on possible selves of the young-old, I would like to take up the influence of media, which Markus and Nurius (1986: 954) list 56 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners 11 In a similar vein, Higgins (1997: 1281 f.) described in his article on the regulatory focus theory that ought selves have a prevention focus whereas ideal selves have a promotion focus. as one example. The representation of aged people in the media - e.g. adverti‐ sements, commercials, and films - has changed considerably. Firstly, the theme of ageing is addressed and older people are shown more often than 20 years ago. Secondly, the way older people are portrayed has changed from being ‘old and feeble’ to a more active, still younger looking individual not to be underesti‐ mated (see chapter 2). This image promoted by the media in recent years has added to the possible selves of the current cohort of the young-old. The goal of staying younger, more active, and healthier for longer thus trickles down to individuals and has the potential to shape their possible selves along with more immediate socio-cultural factors such as significant others. The factor of significant others - family, peer group, teachers, co-workers etc. - especially comes into play with another type of possible self (apart from, for example, ideal selves): ought selves. Ought selves are what we think others expect us to become. But again, as pointed out in section 3.2.1, it is about what we believe others may or may not expect us to be or to become. Like ideal selves, ought selves can act as guides for young-old language learners. The difference between the two is that ideal selves are more related to language learners’ own goals and hopes, whereas ought selves are related to the expectations of significant others (e.g. their partners, children, grandchildren, or friends) regarding an ideal future self for the young-old language learners. Ushioda and Dörnyei (2011: 82) conc‐ lude that ought selves “may therefore bear little resemblance to one’s own de‐ sires or wishes or the possibility of ever attaining them.” 11 On a final note, with regard to ageing and possible selves, crucial life transi‐ tions such as retirement once more come into play (see chapter 2). Cavanaugh and Blanchard-Fields (2002: 355) write that possible selves “facilitate adaptation to new roles across the life span [and] offer a way to bridge the experience of the current self and our imagined future self.” Thus, connecting the research on possible selves to this study’s young-old participants leads to the question in what ways L2 possible selves change or adapt to new circumstances as they grow older. Ryff (1991), for instance, showed in her study that actual and ideal selves approximate with old age (in contrast to younger age groups she tested in her study) and thus described the results of her study as “shifting self-eva‐ luative horizons” over the life-span (294). She concluded from the results of her study in view of the older respondents that there “appears to be a later life gain wherein the ideal self better fits the real self, warts and all, with whom one has become an accustomed traveller.” (ibid.) In other words, the results of this study 57 3.3 The Temporal Self 12 Note that the process of internalization here is linked to the dynamics of selfconcept and the idea of a “working self-concept“ discussed in section 3.2.3 may shed light on how young-old language learners have adapted (or possibly approximated) their ideal L2 selves to their past and actual L2 selves (e.g. critical incidents in the past or current self-perceptions). 3.3.2 Self-Determination Theory (SDT) Going back to the concept of ought selves, they can become ideal selves when they become internalized by individuals because of group or role conformity (Ryan & Irie 2014: 112 f.; Boyatzis & Akrivou 2006: 628). This process forms the basis of the self-determination theory formulated by Deci and Ryan (1985). Dörnyei and Ushioda (2011: 25) summarize it by writing that “SDT’s central notion of a continuum of self-determination focuses attention on how motiva‐ tion for externally defined goals and behaviours may be socialised and gradually internalised” (see also: Deci & Ryan, 1985: 131 f.; Ushioda 2014). 12 The notion of a continuum, however, makes it difficult to draw a boundary between ideal and ought selves in situ: Is my ideal self for a particular domain in my life truly what I want for myself ? Was it me who generated it, or have I internalized what others (through expectations, pressure or force from peer group, my parents, and so‐ ciety in general) see as an ideal self for me, i.e. the ought self ? Boyatzis and Akrivou (2006: 628) do not regard it as a problem if an ought self is intentionally or consciously internalized into one’s ideal self system. But problems do arise if there is no congruence between the two: To the extent this [the ought self] becomes intentionally integrated into a person’s ideal self, there appears to be no conflict among the various selves. But if they are somewhat different and a person works toward the ought self, at some point in the future, they will awake and feel betrayed, frustrated, and even angry at the time and energy they wasted in pursuit of dreams and expectations that they were never pas‐ sionate about. If we connect this to the discussion about self-concept formation in section 3.2.3, it is debatable whether the problem of incongruent ought and ideal selves affects the young-old as much as it affects younger age groups. As stated in section 3.2.3, high school students were more influenced in their self-concept formation by the appraisals of (adult) socializers than young-old learners, who are not dependent on socializers to that extent anymore (see also: Kormos & Csizér 2008: 347). Are the sources for ought selves for the young-old different to those of 58 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners 13 In contrast to Markus and Nurius (1986), who emphasize a multitude of possible selves, Higgins (1987) uses the singular form of ideal, ought, and actual self. younger age groups? And do the young-old internalize external projections at all? 3.3.3 Self-Discrepancy Theory The interplay of ideal and ought selves described in self-determination theory - the degree of internalisation on a continuum of external control and internal regulation - has also been the focus of the self-discrepancy theory introduced by Higgins (1987), however, with the inclusion of the actual self. 13 Higgins (1987: 320 f.) refers to ideal, ought, and actual self as domains of the self. He adds to this dimension a second cognitive dimension: standpoints on the self (own and other). Standpoints on the Self Domains on the Self Actual Ideal Ought Own actual/ own ideal/ own ought/ own Other actual/ other ideal/ other ought/ other Self-Concept Self-Guides Table 3.1 Self-State Representations (after Higgins 1987) The domains of actual, ideal, and ought combined with the standpoints of own/ other create six different types of “self-state representations” (p. 321; see table 3.1): actual/ own, actual/ other, ideal/ own, ideal/ other, ought/ own, and ought/ other (see also Taylor 2011). He points out that both types, actual/ other and particularly actual/ own, constitute a person’s self-concept, whereas the other four types (ideal/ own, ideal/ other, ought/ own, and ought/ other) can be regarded as self-guides. Similarly, Dörnyei - as a foundation for his L2 Motivational Self-System (L2MSS) - has described ideal and ought-to L2 selves as future self-guides in order to highlight this particular temporal facet of the construct. As with possible selves, it is also important to note that self-guides - i.e. the existence or degree of manifestation of ought and ideal self-guides in a person’s self-system - differ between individuals: “[P]eople differ to which self-guide they are especially motivated to meet” (Higgins 1987: 321). In other words, one person has more ought self-guides than another person, who only or mainly has ideal self-guides. Higgins’ self-discrepancy theory is essentially concerned with 59 3.3 The Temporal Self the discrepancy of self-concept and self-guides and the emotional problems this discrepancy may have on the individual: “[W]e are motivated to reach a condi‐ tion where our self-concept matches our personally relevant self-guides” (p. 321). Higgins (1987: 336) investigated and validated two types of emotional problems caused by discrepancies between self-concept/ ideal self-guides (leading to dejection-related emotions) and self-concept/ ought self-guides (leading to agitation-related emotions) in his study. How then can the self-discrepancy theory be applied to young-old language learners? If we believe what this theory postulates - the discrepancies and their emotional effects, respectively - we can assume that a young-old learner who evaluates or realizes today that he or she has not met his or her ideal L2 self he or she hoped for five years ago, might develop a feeling of disappointment or sadness. If he or she perceives in hindsight that his or her self-concept and ought L2 self-guides (i.e. the perception of not meeting the expectations of his or her teacher, partner, children or grandchildren etc.) do not match, he or she might end up with a feeling of fear or tension (Higgins 1987: 336). As we will see in the analysis of the interviews in chapter 6, however, resilience and ageing (see also chapter 2) may influence this interplay of self-guides (Brandstädter & Greve 1994). The degree of self-discrepancy between self-concept and these two types of self-guides also has an impact on his or her future motivation and perfor‐ mance when learning a foreign language. But if we do not view self-discrepancy or self-guides from a present perspective into the past, as in this example, but rather as powerful and future-directed motivational tools, it is worthwhile con‐ sidering Dörnyei’s L2MSS at this point, and with it, the role of self-concept in SLA research. 3.4 Self-Concept in SLA Research: the L2 Motivational Self-System (L2MSS) Sometimes I can’t figure designers out. It’s as if they flunked human anatomy. - Erma Bombeck - To anyone who has ever delved - even just a little - into the role of the self and self-concept in SLA research, it may come as no surprise if I write here that the issue of the self in language learning processes has been shifting more and more to the centre of attention again. I am writing ‘again’ here because, as Pajares and Schunk (2002: 4) point out in their historical overview, “interest in the self and self-beliefs in psychological and educational research has waxed and 60 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners waned”, and after a longer period of neglect, studying the self in educational contexts has started becoming en vogue again ever since the 1980s (ibid.: 13). Like designs in the fashion industry and all new up-and-coming trends, new strategies and models are developed by different researchers to approach a par‐ ticular phenomenon (self-concept) or targeted populations (young-old language learners) to better understand the self in language learning contexts. But as we will see in this section and particularly later in the methodological section, just as skinny jeans do not work on everybody (or every body), some SLA models and research methodologies do not work for all contexts and targeted popula‐ tions (McEwon et al. 2014). From the plethora of SLA studies and theories on the self, I have already addressed some that have been applied to research on the self in educational contexts on a more general level (e.g. Self-Determination Theory by Deci & Ryan 1985 - see section 3.3.2), as well as models or approaches that focus more specifically on the self and language learning (e.g. Network Model by Mercer, 2011 - see section 3.2.2). In this section, I would like to focus on one particular theory: the L2 Motivational Self-System (Dörnyei 2005, 2009). I chose this model over others because of its relation to the temporal facets of self outlined above, helping to shed light on the past, present, and future possible selves of young-old language learners. 3.4.1 L2MSS: Roots and Application Considering the previously discussed focus on possible selves or temporal as‐ pects of self-concept, the L2 Motivational Self-System of Dörnyei fits well with my particular research context and young-old language learners as a targeted population. His theory contains three different components (Dörnyei 2009: 29), two of which I have discussed in section 3.3.3 regarding their psychological roots: Ideal L2 self: Similar to Mercer’s network model (see section 3.2.2.), which contained different L2-related self-concept facets, Dörnyei presents the Ideal L2 self as “the L2-specific facet of one’s ‘ideal self ’”. It is a future self-guide of what an L2 learner would like to achieve in the realm of language learning (promotion focus); Ought-to L2 Self: A future self-guide of what an L2 learner thinks or believes he or she is expected to achieve in the realm of language learning: “This dimen‐ sion corresponds to Higgins’ ought self and thus to the more extrinsic (i.e. less internalised) types of instrumental motives.” (Dörnyei 2009: 29); L2 Learning Experience: motivating factors related to the learning experience but also to the learning environment. This can be the teacher, other young-old 61 3.4 Self-Concept in SLA Research: the L2 Motivational Self-System (L2MSS) learners, the curriculum, and the language learning experience itself (enjoyment or dislike; success or failure). Dörnyei (2009: 29; Dörnyei & Ushioda 2011: 79 f.) based his theory on SLA research (Gardner & Lambert, 1959: integrativeness or integrative motivation) and psychological research (Markus and Nurius 1987: possible selves). Out of a “growing dissatisfaction with integrative motive” (2009: 22) or integrativeness, which had been introduced by Gardner and Lambert in 1959, Dörnyei combined integrativeness with possible selves research. The original idea behind integra‐ tiveness was the wish to get closer to an L2 community by learning the L2 (Masogret & Gardner 2003: 126; Gardner et al. 2004; Ushioda & Dörnyei 2009). This, however, is difficult to apply to many L2 contexts nowadays (Coetzee-Van Rooy 2006). For instance, most young-old learners learning English at adult education centres have very limited or no contact with L2 speakers. Can integ‐ rativeness, in this sense, thus explain their motivation to learn a foreign lang‐ uage or predict their success in doing so at all? Additionally to this, Dörnyei (2009: 24) points out that in the days of an increasingly more important global English, integrative motivation towards the L2 community is losing its relev‐ ance: The language of this global identity is English, and from this perspective it is not at all clear who EFL (English as a foreign language) learners believe the ‘owner’ of their L2 is. This lack of a specific target L2 community, in turn, undermines Gardner’s theoretical concept of integrativeness: in Gardner’s (2001) definition cited above, for example, what exactly would be - to quote Gardner (2001) - ‘the other language community’ that the learner would want to ‘get closer to’? Instead of focusing on an L2 community then, Dörnyei decided to gear integ‐ rativeness towards possible selves. As already mentioned above, he focused on a combination of L2 motivation theory and the psychological concept of possible selves (2009: 24). He postulates that what keeps us motivated to learn an L2 is moving from our actual self towards the future self-image (particularly the ideal L2 self) of successful communication in the L2. With this, Dörnyei has joined the move from target group and environment as key factors for motivation re‐ search in SLA towards more internal aspects of motivation such as self and identity as described at the beginning of this section (Mercer 2011: 30; Dörnyei & Ushioda 2009). It is mainly for the ideal L2 selves, that he developed strategic implications for a “motivational intervention” (2009: 33). He derived these implications from a list of conditions for the ideal L2 self being motivationally effective: 62 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners 14 See also: Erikson & Erikson 2007 1. The learner has a desired L2 ideal self or future self-image; 2. This future self-image is elaborate and vivid; 14 3. It is perceived as plausible and is in harmony/ does not conflict with ex‐ pectations of significant others or other external factors; 4. The desired L2 ideal self is regularly activated in a learner’s working self-concept; 5. It needs “relevant and effective procedural strategies that act as a roadmap towards the goal”; and, 6. The latter also includes negative consequences (feared self) if the L2 ideal self is not achieved. (adapted from: Dörnyei 2009: 32 and Dörnyei & Ushioda 2011: 83 f.) For transferring these conditions into practical implications, the roles of imagery and vision are of particular importance to Dörnyei since their motivational power has been demonstrated in other contexts such as professional sports (Dörnyei 2009: 25). I will not discuss his detailed list of how vision and imagery make learners aware of their ideal L2 self nor how it helps to nurture and main‐ tain it at this point, but rather come back to it during the discussion of my research results and conclusion (see chapters 7 and 8). But can this model be applied to my research context and, more specifically, to the young-old language learners at all? The interplay of language learners’ ideal L2 selves, ought-to L2 selves and language learning experience presented in the framework of L2MSS nicely captures how the temporal L2 selves - espe‐ cially the future-related ones - can influence a young-old’s language learning experience and motivation. However, the past or actual dimensions also need to be considered and included in the L2MSS framework. As mentioned in section 3.3.1, future-related possible selves (ideal and ought-to L2 selves) are not created in a vacuum but have their roots in a learner’s past and actual selves, as Dörnyei acknowledges in his practical implications for creating or constructing the ideal L2 self: Strictly speaking, the term ‘constructing’ the Ideal L2 Self is not really accurate because it is highly unlikely that any motivational intervention will lead a student to generate an ideal self out of nothing - the realistic process is more likely to involve awareness raising and guided selection from multiple aspirations, dreams, desires etc. that the students have already entertained in the past. (2009: 33) Taylor (2011: 56 f.) adds to this criticism that Dörnyei’s L2MSS model is lacking an actual L2 self position from which learners can work towards their ideal L2 63 3.4 Self-Concept in SLA Research: the L2 Motivational Self-System (L2MSS) self. I would argue that the model’s constituent “L2 learning experience”, which has been criticized as structurally different from the other two constituents by Dörnyei (2009: 29) himself, encompasses actual and past L2 selves. As with the effects of the ought-to L2 self, results of my collected data (the interviews in chapter 6 in particular) may offer grounds for an extension of the existing model regarding these past and actual L2 selves that most likely form the up until now broad L2MSS constituent “L2 learning experience”. More importantly, the results of this study may further shed light on the interplay between the two self-guides and the learning experience. Since the age of participants has been an issue with the use or application of research instruments such as questionnaires in other contexts (see end of section 3.2.3), we also need to ask if the L2MSS is suitable for doing research on young-old language learners. McEwon et al. (2014: 32), for instance, notice in their detailed comparison of studies on the Socio-Educational model (Gardner 1985, 2010), Self-Determination theory (Deci & Ryan 1985) and Dörnyei’s L2MSS (2009) that the L2MSS has to date been more often used in research with young learners (<18 years of age, elementary level and secondary level learners) and additional research is required in this area across age groups. In their study on age-related differences in motivation (in which they also tested the two key factors ideal and ought-to L2 self of Dörnyei’s L2MSS model), Kormos and Csizér (2008: 349) came to the conclusion that regarding age it is impossible to find a universal motivation theory, since “it is not only the case that a fixed set of factors play different roles in L2 motivation at different ages but also that certain factors are not even meaningful in a particular setting or for a specific age group.” McEwon et al. (2014) add to this finding by Kormos and Csizér (2008) that some researchers regard the L2MSS as more appropriate for older age groups (in the case of McEwon et al.’s overview, they listed studies with a focus on post-secondary language learners which were 18 years and over, as well as tea‐ chers) because “an individual’s capacity to think self-reflectively and to envision an ideal self might emerge in adolescence and young adulthood” (2014: 34). Others, however, envisage better possibilities for using this model to investigate and understand younger learners because the L2MSS model has - according to McEwon et al. - a strong focus on intrinsic motivation, which may be “parti‐ cularly relevant to younger learners” (ibid.). It becomes clear, with regard to the latter argument against the use of L2MSS with post-secondary language learners over 18 years, that McEwon et al.’s (2014) overview of studies reflects the lack of a more differentiated view of “older” or adult language learners. Indeed, young-old learners are a more suitable target population for the use of the L2MSS model because of their increased capacity 64 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners to think self-reflectively. But even when it comes to the degree of intrinsic mo‐ tivation when starting to learn a foreign language, it is worthwhile differentia‐ ting young-old learners from other adult age groups. As they are moving to‐ wards retirement or are already retired, they have more opportunities to do what intrinsically drives them. Instead of learning or refining their English language skills because they are required to do so for their job, they are free to do so for completely different reasons which are more in line with what they want for themselves (e.g. travelling, meeting other people in class or abroad, or keeping their mind active: see chapters 5 and 6). If these two dimensions - degrees of self-reflectivity and intrinsic motivation - are prerequisites for the application of the L2MSS model, young-old language learners are - of all age groups - the most suitable target population. 3.4.2 Summary At the beginning of outlining my theoretical foundations, I have formed a focus on self-concept - what it is and what it is not - and then moved on to its temporal facets in the shape of possible selves. In the last part of this chapter, I have arrived at a theory - the L2 Motivational Self-System by Dörnyei (2005, 2009) - that takes up my initial theoretical considerations of self-concept and that is directed more specifically at SLA research than most other existing self-models and the‐ ories. Yet, going back to the fashion metaphor at the beginning of this section, the skinny jeans might still need mending and adapting to fit my research per‐ fectly. What does this mending process entail? In terms of age, the young-old might seem to constitute a suitable target group for the application of the L2MSS, but other facets of the temporal or possible selves theory might have to be added in order to fully comprehend the language learning process of young-old lear‐ ners. To grasp the full picture of young-old language learners’ possible selves, other temporal dimensions, namely past and actual selves, need to be explored since future-related ideal selves and ought-to L2-selves build on them. 65 3.4 Self-Concept in SLA Research: the L2 Motivational Self-System (L2MSS) 3.5 Research Questions Using traveling as a metaphor, a research question is the destination of your journey, and where you choose to go to some extent determines your means of transportation, what you will bring on the journey, who you would like to accompany you, etc. - Nugrahenny 2012: 47 - Research questions constitute the focal point of any research project, and finding answers to them shapes and guides the whole research process and its metho‐ dological design (methods, instruments, data analysis). My research questions are anchored in the theoretical areas I have outlined in this and the previous chapter. But it may also be accurate to say that they have grown out of the theoretical foundations, since research questions can - especially in qualitative research - change, develop, or evolve over the course of a research project (Nu‐ grahenny 2012: 48; Maxwell 2005). Furthermore, Nunan (1992: 213) provides preconditions for good research questions by pointing out that they should be “worth asking in the first place” and be “capable of being answered.” What does the development of my research questions look like and to what extent do they comply with what Nunan posited as preconditions of good re‐ search questions? Looking back to the beginning of my research journey, I started out with the following research questions: How do older language learners (60-75) learn foreign languages in a formal lang‐ uage learning environment? Are older learners (60-75) affected by… …their life experience… …their self-image (age, culture)… …when learning a foreign language? On the basis of the answers to the questions above: What does this mean for the development of language learning material for older learners? These initial questions are worth asking, but are not amenable to answers - at least not within the given time frame and resources available to me. The ques‐ tions are very broad and difficult to grasp or answer (e.g. how to measure life experience? ) and are in some ways limited to the wrong areas (a strong focus on chronological age only). Even though this research project will undoubtedly point to necessary changes in the design of material for young-old language learners, it is probably too broad, and thus probably misdirected for research purposes, to include the third question above as a research question. Therefore, I have considerably restructured and limited my research questions and research focus to the following: 66 3. Self-Concept, Its Temporal Facets, and Foreign Language Learners How do the young-old conceptualize themselves as foreign language learners? How do they perceive their past, present, and future L2-self/ selves? Which factors (e.g. age-related: retirement) influence their language learner self-concept? My main question is thus now followed by two sub-questions. In contrast to the set of research questions I started off with, these questions show a narrower focus on the temporal aspect of self-concept, i.e. possible selves. I have moved away from using chronological age to define the target group and have instead adopted the term young-old which encompasses the different dimensions of age and ageing (see chapter 2). What are the best tools to tackle my new research questions? As I have men‐ tioned earlier in this section, research questions guide the researcher and shape the methodological framework or design. And, as discussed in the previous chapter as well, the methodological framework needs to fit the individual re‐ search project just like a piece of clothing fits a particular body. In view of the complexity of investigating the young-old language learner’s self-concept, a mixed method design has been chosen to address the research problem. In the next chapter, I will explain why this methodological framework is suitable for my research and outline the tools within my mixed methodology design, to help find answers to my research questions. 67 3.5 Research Questions 4. Methodology, Research Design, and Research Context The focus of this chapter is on a detailed description and discussion of the me‐ thodological blueprint of my research project. Using Creswell ’s (2009: 204 f.) checklist and guiding questions for designing mixed methods research as well as the article on publishing mixed method research (hereafter MMR) by Mertens (2010) as reference points for the structure of this chapter, I focus on aspects such as definitions of mixed methods research, its background and characteris‐ tics, reasons for employing mixed methods and what to consider when em‐ ploying them (degree of interaction, timing, prioritizing/ weighting, and mixing). Following this, I will briefly present the quantitative and qualitative research instruments used for my mix and discuss in what way they relate to the MMR design, criteria, and issues. In the final section of this chapter, I will introduce the research context of “vhs centres” (German adult education centres). They constitute and provide the learning environment of the young-old language le‐ arners in this study. As will become apparent in the following paragraphs, MMR is a fairly new methodology next to the two more established qualitative and quantitative me‐ thodologies. Even though the combined use of quantitative and qualitative data can be traced back to a study conducted by Campbell and Fiske (1959), its ma‐ nifestation - e.g. in the shape of an MMR-specific terminology - can still be considered in its adolescence. The recently skyrocketing number of studies em‐ ploying this approach (Bergman 2008: 1; Tashakkori & Teddlie 2010: 809 f.; Morse 2010: 340; Creswell & Plano Clark 2007: 14) is another indicator that it is evolving and forming a distinct shape more rapidly than ever before. Its recent develop‐ ment has seen a heated debate about its characteristics, conventions, and quality criteria. I hope to apply the insights from these new developments in MMR to the process of my own research in the following. I will outline how I have ad‐ dressed the MMR challenges to pursue my research in the field of foreign lang‐ uage teaching and learning as well as to find answers to my research questions. In addition to this, all of the emerging issues of MMR in research as a whole are even younger in nature in the field of foreign language teaching and learning research. Given this research field’s traditionally stronger qualitative tradition (Schramm: 2016; Dörnyei 2007: 187; Macintyre et al. 2009: 53), MMR still needs further development and refining to be employed in the language learning con‐ text. Researchers and practitioners can profit from the combined use of quan‐ titative and qualitative strands in order to better understand the complexity of the language learning classroom (Knorr & Schramm 2016: 90; Turner & Meyer 2000; Dörnyei 2007: 313 f.). In their review of language teaching and learning studies employing MMR methodology, Riazi and Candlin (2014) found that there is still a lack of discussion of theoretical constructs (see chapters 5 & 6) and the purposes of MMR (see sections 4.1.2 & 4.2), in order to produce coherent research and valid outcomes. They conclude that foreign language teaching and learning researchers need to become familiar with the rules of this methodology and take a critical stance. However, facing a “dearth of extensive critical discussion on MMR in language teaching and learning” (ibid. 161), they admit that this is a complex task. 4.1 Mixed Methods Research - Why? , What? , How? Writing about mixed methods and the specific design is important because it foreshadows the methods and provides a means by which readers can evaluate the study. - Creswell & Plano Clark 2007: 89 - As mentioned above, one sign of MMR’s currently evolving nature is in terms of its terminology and this, as Bryman (2008: 88) and Johnson et al. (2007) point out, is not without its problems. It starts with the name: researchers have used different titles or key words for this way of doing research such as combining, integrating, hybrid, methodological triangulation, multitrait method, blending, meshing, or merging (Creswell & Plano Clark 2007: 5 f.; Bryman 2008). An ag‐ reement has been established for differentiating mixed methods from multiple or quasi-mixed methods. The different names indicate whether quantitative and qualitative data are fully integrated, i.e. if we are dealing with one complete study (mixed methods), or if there remains separation of the two types of data and if we are thus dealing with two studies that could stand alone (multiple or quasi-mixed methods; Morse & Maddox 2013: 224 f.). Thus, to emphasize that I aim to integrate both qualitative and quantitative data in a single study but also to join the growing number of researchers using this terminology (Creswell & Plano Clark 2007: 6), I will refer to this approach as mixed methods research (MMR) in the context of my research. I align my understanding of MMR with the definition presented by Johnson et al. (2007: 123), which they have synthesized from a range of definitions from leading researchers in the MMR-field: 70 4. Methodology, Research Design, and Research Context Mixed methods research is the type of research in which a researcher or team of researchers combines elements of qualitative and quantitative research approaches (e.g., use of qualitative and quantitative viewpoints, data collection, analysis, inference techniques) for the broad purposes of breadth and depth of understanding and cor‐ roboration. With this definition, Johnson et al. cover the key questions guiding the following sections in this chapter: Why? (not only the broad but also more specific pur‐ poses of MMR will be looked at); What? (the qualitative and quantitative ele‐ ments making up this MMR study); and, How? (the design and process of data collection, analysis and inference making). Adding to the definition by Johnson et al. (2007), I share Bryman’s (2008: 89) view that MMR is more than the mere sum of its quantitative and qualitative parts and that it is this what sets MMR apart from these approaches. Teddlie and Tashakkori (2010: 16 f.) further highlight the distinctiveness of MMR compared to other methodologies with the following two characteristics: a) “Rejection of either-or at all levels of the research process”; and, b) “Subscription to the ite‐ rative, cyclical approach to research.” (see chapter 7) While the latter is strongly facilitated by the actual design of my research project, which will be outlined and discussed in a later section of this chapter, the first characteristic listed here points to the methodological eclecticism un‐ derlying mixed method research. Teddlie and Tashakkori see the MMR resear‐ cher as a “connoisseur of methods” (2010: 16) and bring up Rossman and Wil‐ son’s (1994) statement that MMR researchers are “shamelessly eclectic”. The methodological eclecticism of MMR is aligned with the philosophical orientation of pragmatism, which is regarded as the most common conceptual stance among MMR studies (Creswell & Plano Clark 2007: 23) and which is also part of the foundation of my research project. Nevertheless, there are also MMR studies - given they view MMR as a method instead of a methodology, and depending on the weighting of quantitative and qualitative components (see section 4.1.3) - that rely on either postpositivism or constructivism (ibid. 27 f.). Going back to pragmatism, Creswell and Plano Clark (2007: 23) define it in relation to MMR as follows: The focus is on the consequences of research, on the primary importance of the qu‐ estion asked rather than the methods, and multiple methods of data collection inform the problems under the study. Thus it is pluralistic and oriented toward ‘what works’ and practice. Due to the fact that the pragmatic worldview is oriented toward “what works” and a “practical problem-solving attitude”, McCaslin (2008: 9) argues that studies 71 4.1 Mixed Methods Research - Why? , What? , How? 1 Related to this, Riazi and Candlin (2014: 167) see a developmental phase of ‘principled eclecticism’ in the use of MMR in language teaching and learning research. with this conceptual stance are less prone to the effect of ‘When all you have is a hammer, everything looks like a nail.’ - though, on the other hand, this should not seduce researchers into thinking “When in doubt, mix methods…” (Dörnyei 2007: 46) 1 Creswell and Plano Clark (2007: 23) add that, when based on prag‐ matism, MMR can combine deductive thinking (or top-down), which is associ‐ ated with a postpositivist research procedure, and inductive thinking (or bottom-up), which is associated with a constructivist research procedure (see also Tashakkori & Teddlie 2010: 16, for a top-down/ bottom-up distinction of research-driven dimensions in MMR). This ultimately also affects the research question(s) in a research project. 4.1.1 Intermezzo: Research Questions Revisited Like the process of conducting mixed methods research itself, mixed research questions are necessarily complex and dynamic. The field of mixed methods is only beginning to think about the nature of these questions. - Plano Clark & Badiee 2010: 300 - Even though I have already discussed the development of my research questions in relation to the theoretical foundations and research context in the previous chapter, looking at the research questions through an MMR lens here might lead to a further readjustment or refinement of them. The main question arising from the discussion in the previous section is how research questions reflect the combination of inductive/ deductive, top-down/ bottom-up, or qualitative/ quan‐ titative approaches in the way they are formulated and structured. Interestingly, very few researchers have explicitly addressed this issue previously (Creswell & Plano Clark 2011: 161). Just as, for example, the ‘signature’ terminology of MMR is in the process of development, other constituents of this methodology such as the research questions or designs (see section 4.1.4) are also evolving. These signature MMR research questions need to indicate a need for mixing quantitative and qualitative data sets in order to find answers (Teddlie & Tas‐ hakkori 2008). Plano Clark and Badiee (2010) have described four dimensions for the writing of MMR research questions: rhetorical style (format; level of integration), the relationship of questions to other questions; and the relationship of questions to the research process. In what ways are my research questions aligned with these dimensions? 72 4. Methodology, Research Design, and Research Context • Rhetorical style (format) Plano Clark and Badiee list questions, aims, or hypotheses as possible ways to formulate research questions. In the case of my study, research questions come in the form of questions. In the process of data analysis of the quantitative data set, however, hypotheses will be formulated on the basis of the quantitative aspects of the research questions in order to analyse the survey results (see chapter 6). • Rhetorical style (level of integration) The two authors identify five different ways researchers employ or struc‐ ture their questions in order to express the degree of integration of dif‐ ferent data in the design: “Broadly speaking, mixed methods researchers state either separate quantitative and qualitative questions or mixed me‐ thods research questions or both.” (2010: 293) Taking a look at my research questions, it becomes clear that they are a combination of different types of questions: an overarching question calling for an MMR approach, fol‐ lowed by two sub-questions which hint in their formulations at em‐ ploying qualitative or quantitative measures. In what way is the overar‐ ching research question “How do the young-old conceptualize themselves as foreign language learners? ” calling for an MMR approach? Plano Clark and Badiee (2010: 294) report a study in which researchers indicate the justification of an MMR approach by referring to the language used in the question. They argue that the word how, like the formulation of my over‐ arching MMR research question, is indicating a qualitative and the word do a quantitative component (i.e. adding a component similar to a hypo‐ thesis that needs to be confirmed or disconfirmed). Put differently, star‐ ting the question “How do they conceptualize themselves? ” contains two layers: firstly, to find out if they do or don’t conceptualize anything re‐ garding their language learning experience, at all. Secondly, if certain as‐ pects of this can be confirmed, how is their language learning self formed or what shape and values are attached to this? Furthermore, Plano Clark and Badiee report another study where researchers used the word perceive (comparable to my use of conceptualize) and, by treating it neutrally, jus‐ tified an MMR approach because answering this question “required qua‐ litative and quantitative information to adequately describe the percep‐ tion of impact.” (2010: 294). • The relationship of questions to other questions Does one question depend on the findings of another question (depen‐ dent) or are they related but independent questions? In the case of my research questions, I have chosen an independent option for their relati‐ 73 4.1 Mixed Methods Research - Why? , What? , How? onship, which is in line with what MMR researchers have noticed in exis‐ ting studies to be the better combination for convergent designs (Creswell & Plano Clark 2011: 163). In addition to this, in some MMR projects the research questions are ordered according to their importance. Since the quantitative and qualitative strands are treated equally in my design (see section 4.1.4), the order of the sub-questions is irrelevant. • The relationship of questions to the research process Here, the authors differentiate between predetermined and emergent op‐ tions for writing research questions. Predetermined research questions are set at the beginning of a study based on, for example, literature, whe‐ reas emergent questions can be (re-)formulated in the process of data collection, data analysis, or interpretation. Creswell and Plano Clark (2011: 162) see a tendency of predetermined questions being used more often in convergent designs. In my study, however, I would argue that my research questions are emergent in nature since they have been refined beyond the review of the literature (see chapter 3) through other stimuli during the research process (e.g. the data itself or discussions with fellow researchers). Moreover, interesting (research) questions for further re‐ search will eventually emanate from this study’s interpretation stage or findings. The way I have positioned my research questions within these four dimensions relates or foreshadows my MMR design, which will be further outlined in the following. However, as I have mentioned at the beginning of this intermezzo, further refinement of the research questions may be necessary. For this, I would like to take Plano Clark and Badiee’s (2010: 294 f.) suggestion to include a pro‐ cedural/ mixing question (rhetorical style: level of integration) for the purpose of “providing initial direction to a study where topics are expected to emerge, highlighting the use of an integrated design, foreshadowing integration proce‐ dures for audiences unfamiliar with mixed methods, and conveying the com‐ plexity inherent in mixed methods designs.” Adding this extra question results in the following research question framework for my study: How do the young-old conceptualize themselves as foreign language learners? a. How do they perceive their past, present, and future L2-self/ selves? b. Which factors (e.g. age-related: retirement) influence their language learner self-concept? c. In what ways do survey and interview study converge? 74 4. Methodology, Research Design, and Research Context The procedural/ mixing question (c.) further emphasizes the fact that both types of data will be eventually integrated and underlines the use of a convergent type of MMR model (see section 4.1.3). Since this sub-question is a different type of question (methodological), it will be integrated and addressed differently than questions a. and b. in the discussion (chapter 7). 4.1.2 Why? - Reasons for Employing MMR We should be shamelessly eclectic in our use of methods to understand the intractable and persistent problems of education today. - Rossman & Wilson 1994: 326 - Several researchers in the field of MMR have highlighted the importance of giving reasons for mixing methods and being ‘shamelessly eclectic’. Thus, before going into details about what I am going to mix and how quantitative and qua‐ litative parts will be combined in the following sections, I will firstly give reasons for employing this methodology for my research purposes. Bryman (2008) has conducted a comprehensive mixed method study to find out more about the rationales given by researchers to use this way of doing research. He was looking for rationales that go beyond a simple fashion or fad in recent research tradition, and which also point to the idea that, as already mentioned above, MMR is more than the sum of its qualitative and quantitative parts. Drawing from those rati‐ onales I am going to outline my reasons for employing mixed methods in my research project in this section. A first major source for my methodological rationale is a review of the lite‐ rature. As mentioned at the beginning of this chapter, the complexity and multi-layeredness of phenomena in foreign language research - including lang‐ uage learner self-concept - is better understood through the use of both quali‐ tative and quantitative strands (Knorr & Schramm 2016: 90; Candlin & Riazi 2014; Dörnyei 2007: 186 f.; Turner & Meyer 2000). In the case of my research on language learner self-concept, Mercer (2011a: 6, 2011b: 338) has pointed to the majority of quantitative approaches to researching the self or self-concept (par‐ ticularly in psychology; see also Mummendey & Grau 2008; Reinders 2011a). Mercer, thus, has focused her research on doing qualitative research on the language learner self-concept using case studies and Grounded Theory in order to better capture the contextual or situated nature of the language learner self-concept. Ryan and Irie (2014: 120) also see the quantitative approach to studying the language learner self playing a major role but are critical about the applicability of quantitative research results for educators: 75 4.1 Mixed Methods Research - Why? , What? , How? Educators need to know more about the real stories learners tell themselves and how these are incorporated within their self-concept. The extent to which events, actual or imagined, are incorporated into the self-concept is very much an individual sub‐ jective judgment. This suggests that it may be more profitable for researchers to in‐ vestigate this subjective dimension, rather than adhere to notions of objective inves‐ tigation. Apart from the literature review, rationales are closely connected to or derive from research factors such as the research questions, existing research using MMR, and philosophical assumptions underlying the research. As mentioned above, researchers conducting an MMR study tend to neglect bringing up and discussing the reasons for employing this particular methodology. But even if researchers list the rationales in their reports, they often do not match the ra‐ tionales in practice (Bryman 2008; Candlin & Riazi 2014: 160 f.). One of the re‐ asons for this seems to be that these researchers did not expect certain outcomes when starting out with an MMR study or merging the different data. The op‐ posite of this is also the case. Bryman refers to this as the ‘Gatling gun approach’ to rationales for MMR. In the sense of ‘more is better’, some researchers are inclined to make a long list of rationales. He hypothesizes that uncertainty about using this methodology may be a reason for this. Another reason, however, could be that a researcher is well aware of the complexity of research reality. In sum, MMR in particular can result in more outcomes and thus reveal more rationales than anticipated (Bryman 2008: 94 f.). Thinking about and discussing the reasons for choosing an MMR approach over another methodology is creating solid ground for data collection, its ana‐ lysis, and interpretation of the results. Before moving on to explain my actual design, I am going to introduce and discuss the rationales I consider relevant to my use and expectations of MMR in this research project. For this, I draw on Bryman’s study on MMR rationales in theory and in practice, which he in turn derived partly from the five major rationales (triangulation, complementarity, development, initiation, and expansion) extracted from a range of MMR studies by Greene et al. (1989): • Completeness or Expanded Understanding (Creswell 2009: 204 f.) As noted earlier in this section and in the chapter on theoretical founda‐ tions of my research, language learner self-concept and its temporal facets are complex. Mixed methods can provide a better understanding of its complexity than either a qualitative or a quantitative approach could alone (Candlin & Riazi 2014: 156; Dörnyei 2007: 313 f.; Lazaraton 2005: 219; Schramm 2016: 54 ff.). In the words of Creswell and Plano Clark 76 4. Methodology, Research Design, and Research Context (2011: 8): “One type of evidence may not tell the complete story, or the researcher may lack confidence in the ability of one type of evidence to address the problem.” • Offset This rationale is based on the notion that both quantitative and qualitative research have their strengths and weaknesses. By bringing them together, they offset each other in their weaknesses. In the case of my research, I have received interesting answers to the questions in my survey. They have been provided by a high number of participants, reducing the bias and increasing the validity of this part of my study. However, the down‐ side of this is that I cannot go back to the survey participants and ask further questions about interesting or unexpected results. Therefore, the interviews, i.e. the qualitative part of my study, provide the depth the survey lacks. • Utility As mentioned above, Ryan and Irie (2014) called for approaching research on learner self-concept in a way that would better capture its subjective dimension as well as its contextuality and thus improve the applicability or, as Bryman refers to it, utility for practitioners in the field of language learning. Connected to this, Dörnyei (2007: 166) adds that employing a mix of methods helps to make it ‘more palatable’ for a broader audience, i.e. scholars working with either a qualitative or quantitative paradigm. • Validity (see also: Morse & Maddox 2013; Hammersley 2008), Inference Quality (Teddlie & Tashakkori 2010), or Corroboration Finally, combining the results of the qualitative and quantitative strands in my study enhances the validity of the overall interpretation or meta-in‐ ferences (Teddlie & Tashakkori 2008: 101). The term validity, however, is more commonly used in the context of quantitative studies, thus Teddlie and Tashakkori (2008) proposed the term ‘inference quality’. I use this term to further highlight that both approaches are treated or weighed equally in my study. The last rationale might especially raise issues regarding inference quality cri‐ teria. Are the results of the study still valid if there are incongruities between qualitative and quantitative results and inferences? Section 4.2 is dedicated to possible criteria as well as strategies for dealing with incongruent data results. 77 4.1 Mixed Methods Research - Why? , What? , How? 4.1.3 What? - Research Instruments This section explores the quantitative and qualitative strands or components of my study, before explaining in detail how these components will be combined in the next section (4.1.4). In the brief introduction of the instruments it will become clearer in what way the aforementioned rationales come into effect. A more in-depth discussion of the research instruments’ development, contents, and structure, however, will be provided in chapters 5 and 6. The survey represents the quantitative element of my MMR study and has been developed on the basis of self-concept literature in the field and adapted questions from already existing questions by, for example, Dörnyei (2010). After an initial testing of a pilot survey with a follow-up revision of the structure and questions, survey data has been collected from six different Volkshochschulen in the German state Hesse (see section 4.3). Altogether, 195 participants returned their completed questionnaire. Due to time constraints, a preliminary analysis of 50 surveys using t-tests and ANOVA tests (see chapter 5.4) was completed with its results partly contributing to the construction of the interview ques‐ tions. One example of this is the survey question “The way I study English has changed over time.” (From German: “Die Art und Weise, wie ich Englisch lerne, hat sich mit der Zeit verändert.”). The preliminary analysis of the answers to this question showed that a greater proportion of the participants stated that this statement applied to them (Likert scale values 4-6). In this case, the survey’s limits concerning its depth became visible. Thus, the rationale ‘offset’ comes into play, as only the qualitative strand can provide further insight into this topic by asking in what ways exactly participants have changed approaching EFL learning. The interview study was conducted shortly after having received all questi‐ onnaires from the survey study and the preliminary analysis. The questions of the semi-structured interview were mainly based on the literature, but, as out‐ lined in the example above, the preliminary analysis of the survey helped to further accentuate a few topics in the interview. As with the questionnaires, the main interview study was also preceded by a pilot study in order to test the instrument. Four young-old language learners from a language class I taught for one semester as a substitute teacher participated in the pilot interview study. While the 195 participants of the survey study were from other vhs centres in the state and participated anonymously, the participants of the main interview study were 21 young-old learners from three of my own language classes I had been teaching for three to five years (intermediate and upper-intermediate le‐ vels; CEFR-level B1-B2). The transcribed interviews were analysed using a qua‐ litative content analysis (see chapter 6.3) and to avoid a potential conflict bet‐ 78 4. Methodology, Research Design, and Research Context ween my teacher/ researcher role and ensure reliability of the category system, an inter-coder agreement check was carried out (see chapter 6.4). In terms of this and the other rationales mentioned in the previous section, it becomes apparent how both qualitative and quantitative strands in my study have their strengths and weaknesses, and in what ways survey and interview together can produce a more complete picture of language learner self-concept of young-old learners. The quantitative part of my research project balances the weakness of my qualitative study, i.e. the bias caused by my role of researcher and teacher to the interview study participants (offset and inference quality/ validity). The rationale utility is addressed with this combination of data as well. On the one hand, practitioners are able to apply the meta-inferences to a broader population because of the survey representing young-old learners from rural and urban vhs centres in the state of Hesse as well as from a range of English class levels. On the other hand, English teachers also get a valuable in-depth perspective on the complexities of the young-old language learner self through the qualitative strand in my research project. 4.1.4 How? - Finding the Right Mix In the following, I am going to present the structure of my MMR design using the common models and the four criteria of level of interaction, priority (or weighting), timing, and mixing (Creswell & Plano Clark 2011: 64-68; see also: Creswell 2009; Creswell & Ivankova 2009) as reference points. In an earlier ver‐ sion of these four criteria, Creswell and Plano Clark (2007) have listed the factor of ‘theorizing’ instead of the level of interaction between qualitative and quan‐ titative strands. Since the theoretical and conceptual stances have already been outlined above, this criterion will not be considered again here. It will become clear as this section unfolds how some of the rationales discussed in section 4.1.2 influence the structure of my MMR design. Firstly, Creswell and Plano Clark (2011; with reference to Greene 2007) point out that the level of interaction between the quantitative and qualitative com‐ ponents needs to be considered when choosing a MMR model or design. They differentiate between independent and interactive levels. In the independent option both strands are processed separately and only mixed at the final stage of interpretation. In the interactive option, the two methods are mixed at some point(s) before the final interpretation. Referring to my design outlined in the previous section on what I have collected and analysed in my research project, one could criticize the authors’ presentation of this criterion. In my design, a direct mixing of the two strands before the interpretation stage involved the 79 4.1 Mixed Methods Research - Why? , What? , How? adaptation of only a few interview questions; the majority of the interview guide is, however, based on theoretical input and research questions. This shows that the independent/ interactive distinction is not necessarily a black and white de‐ cision, but should rather be placed on a scale from fully integrated/ interactive to fully independent. On this scale, my research design would be placed closer to the fully independent side. The next factor for deciding on an MMR design is about the degree of priority given to each of the strands. This criterion was referred to as “weighting” in an earlier edition of their book (Creswell & Plano Clark 2007). In the notation system specific to the emerging MMR terminology, the prioritizing of one strand over the other is expressed by the use of capitalization (Teddlie & Tashakkori 2010: 11 f.; Morse 2010; Creswell 2010: 57). For instance, QUAL/ quan would in‐ dicate that priority is given to the qualitative part of an MMR study with the quantitative part being supplemental (Morse 2010: 340 f.). The use of QUAN/ QUAL then indicates that equal weight is given to both strands, as is the case in my study: the results of the interview and the survey are regarded as equally valuable in my study and contribute to eventual meta-inferences to the same extent. In terms of timing, Creswell (2009) categorized different MMR models into sequential and concurrent designs. In the second edition of their handbook, Creswell and Plano Clark (2011: 66) add a third timing possibility to this list: ‘multiphase combination timing’ containing three or more phases with a possible combination of sequential and concurrent timing. The interesting qu‐ estion about the timing factor is which stage(s) need to happen in a particular order for a design to be classified as a sequential or a concurrent design. Taking a look at my design, one could argue that I have conducted a sequential design since collection of quantitative and qualitative data happened one after the other. Creswell and Plano Clark (2011: 65), however, stress that the criterion timing refers to “the order in which the researchers use the results from the two sets of data within a study - that is, timing relates to the entire quantitative and qualitative strands, not just data collection.” Thus, a classical example of a se‐ quential design would include a phase in which quantitative data is collected and completely analysed first, then qualitative data is collected as a follow-up phase, in order to further explain quantitative results (as in the prototypical explanatory sequential design; Creswell & Plano Clark 2011: 69). In this case, the more crucial phases of analysis and points of time of interaction (or merging of interpretations) between the strands are sequential in nature. In the case of my study, the qualitative component was not conducted for the sole purpose of explaining the quantitative results. At that stage, the preliminary results of the 80 4. Methodology, Research Design, and Research Context survey would not have justified or provided enough information for conducting a supplemental interview study. Even though data collection happened in a more sequential way - which is also explained by the fact that data collection was carried out by one researcher and not a research team - the crucial phases of my study (data analysis and interpretation) give it a stronger concurrent timing characteristic. The fourth criterion that directs the decision for my MMR design concerns the points of interface (Where are both strands merged? ; Creswell & Plano Clark 2011: 66) and the procedures or ways of mixing qualitative and quantitative data. As pointed out in the previous section, some interrelation between my survey and the interview study occurred at the instrument design stage of the interview questions. Furthermore, both quantitative and qualitative strands build on the same theoretical framework and include or are structured by the temporal as‐ pects of language learner self-concept (for a detailed description see chapters 5 and 6). Creswell and Plano Clark (2011: 67 f.) refer to this as theoretical frame‐ work-based mixing. The crucial mixing or merging in my design, however, is done in the final interpretation stage of the research process after both data sets have been analysed. This decision also has an impact on the writing up stage of the research project and its results (Dahlberg et al. 2010). Analysis of both data sets are carried out separately and presented in two separate chapters (5 and 6). Using SPSS, the analysis of the quantitative data focuses on finding answers or testing hypotheses derived from the research questions. The qualitative data is analysed using a qualitative content analysis method to get a general sense of the topics addressed in the interview regarding the temporal aspects of language learner self-concept (Bryman 2008: 89). In chapter 7, the interpretation and dis‐ cussion (i.e. the mixing or merging of the two strands) of the results of the two analyses are presented. Different combinations of the decisions over these four criteria have led to the development of different models or typologies for MMR designs (Creswell & Plano Clark 2011: 69 f.; Creswell et al. 2008; Nastasi et al. 2010). Thus, the way I have positioned myself in view of the four factors, my design (figure 4.1 below) comes closest to what Creswell and Plano Clark have called the convergent pa‐ rallel design (2011: 69; figure 4.2 below). In their earlier 2007 edition, they called it ‘triangulation design’ but later changed its name to avoid confusion with the term ‘triangulation’ also used in qualitative research (Creswell & Plano Clark 2011: 77; Bryman 2008: 88). This move shows once more how the MMR termi‐ nology is evolving, as mentioned in the introduction to this chapter. The differences between my design (figure 4.1) and the convergent parallel design ‘template’ (figure 4.2) reflect the discussion in the previous paragraphs 81 4.1 Mixed Methods Research - Why? , What? , How? and illustrate in what ways I have deviated, based on my positioning, from the original typology. Does this make my research design invalid? Not necessarily. In fact, researchers such as Teddlie and Tashakkori (2010: 24 f.) and Creswell and Plano Clark (2011) caution that researchers interested in employing mixed me‐ thods should not necessarily view these models or typologies as a “cookbook recipe” (Creswell and Plano Clark 2011: 60). Rather, in the tradition of pragma‐ tism, they should use them as a guide to choosing, from the variations available, the most suitable combination or elements to address research problems and questions: “Typologies provide the researcher with a range of available options to consider that are well defined, facilitate the researcher’s use of a solid ap‐ proach for addressing the research problem, and help the researcher anticipate and resolve challenging issues.” (ibid.) In connection with how my research design deviated from some of the pro‐ posed criteria and models in theory, I would like to add to their advice that in ‘research reality’ things do not always go as planned. The researcher’s degree of experience and skills (Creswell & Plano Clark 2011: 13 ff.) regarding the qua‐ litative and quantitative approaches as well as the timing or planning can affect a research process. Moreover, the availability of participants and the cooperation with institutions and teachers can also influence the actual research process. The next section will address some other unexpected outcomes when it comes to the interpretation stage of the MMR process and discuss if they might be what could be called a ‘blessing in disguise’. 82 4. Methodology, Research Design, and Research Context QUAN (survey) Data Collection Data Analysis Data Analysis QUAL (interviews) Data Collection Data Analysis (preliminary) results compared QUAN + QUAL interpretation and informing RQ Figure 4.1 My MMR Research Design based on the Parallel Convergent Design 83 4.1 Mixed Methods Research - Why? , What? , How? Quantitative Data Collection and Analysis Qualitative Data Collection and Analysis interpretation compare or relate Figure 4.2 Convergent Parallel Design (Creswell & Plano Clark 2011: 69) 4.2 Agree to Disagree? Inference Quality and Handling Divergence [T]he outcomes of mixed methods research are not always predictable. In fact, it may be more accurate to say that they are rarely predictable. - Bryman 2008: 95 - Once the why, what, and how of MMR in my research study have been set and data analysis and interpretation are in progress, it happens that the different data strands produce divergent results - even if the researcher has taken care to ensure that both strands address the same theoretical construct (Candlin & Riazi 2014: 154). So far, very little has been discussed with regard to ensuring the quality of results and dealing with divergence. This may be due to the un‐ derlying assumption that the data analysis and interpretation stages are sup‐ posed to be a harmonious act, otherwise indicating a failure of an MMR design. Bringing up again the term ‘inference quality’ (also one of the rationales for this study discussed in section 4.1.2), Teddlie and Tashakkori (2008) suggest a fra‐ mework for evaluating the quality of meta-inferences (i.e. an overall conclusion or understanding) drawn from data analyses and the interpretation of qualitative and quantitative data sets. The possible challenge when assessing the quality of inferences made in MMR is that researchers using this methodology need to comply with the standards of three different domains of their design: quantita‐ tive standards, qualitative standards, and meta-inferences (MMR) based on both these strands (Teddlie & Tashakkori 2008: 110). They point out that “[t]his is 84 4. Methodology, Research Design, and Research Context especially difficult when the two sets of inferences are not consistent with each other.” (ibid.) What happens when both data sets are merged or integrated in the final stage and meta-inferences are made? The example of the survey question and its pre‐ liminary analysis from section 4.1.3 serves as a possible scenario to illustrate the difficulty mentioned in the quote above and also why the final interpretation stage is the most crucial - and possibly the least predictable - phase of my MMR design: While a high number of participants in the survey study agreed with the statement that they have changed their ways of studying English over time, an initial analysis of the interviews has not confirmed or supported this finding. Most interview study participants answered this question with the answer “no” and/ or could not think of any particular aspects of their English learning (e.g. strategies) they had changed over time. Here, bringing the two analyses together in a final interpretation yields divergence in the findings. The question of “What went wrong? ” may arise at this point. Addressing possible answers to this qu‐ estion, Creswell et al. (2008: 72) write in a way that is both worryand hope-in‐ ducing: Disconfirming findings may indicate flaws or inconsistencies in the research design. Erzberger and Kelle (2003) explain that discrepancies between quantitative and qua‐ litative data may be the result of researcher’s errors in data collection and analysis or poor application of theoretical propositions. Divergent findings, however, can also be thought of as a means to uncovering new theories or extending theories. Is it errors in earlier stages of the research process (e.g. sampling, data analysis, appropriateness of the chosen MMR model for the research problem) that caused the divergence? Or is it rather a revelation of something new? Is it, for example, a new perspective on the self-concept in young-old language learners? Or, to be more precise, does the inconsistency show two perspectives on one pheno‐ menon - an argument that is also supported by Teddlie and Tashakkori who state that “[i]ntegration does not necessarily mean creating a single understan‐ ding on the basis of the results” (2008: 115; my italics). Concerning the two scenarios of convergence and divergence in findings, Creswell et al. (2008: 70 f.) present an overview of issues in conducting MMR and present strategies collected from the literature for dealing with arising issues for concurrent, sequential, and all designs. For the problem “contradictory fin‐ dings” they list the following strategies: • Help uncover new theories • Collect additional data • Reanalyse original data 85 4.2 Agree to Disagree? Inference Quality and Handling Divergence • Use as a springboard for new inquiry (see chapter 8) • Give priority to one form of data From this list, the reanalysis of data sets and the formulation or uncovering of new theories were used as steps for explaining or dealing with divergence in the interpretation stage of my study. Similar to the ones listed by Creswell et al. (2008) above, Teddlie and Tashakkori (2008: 116) offer more detailed steps in the case of disconfirming findings in the context of their quality framework for MMR inferences (quality criterion ‘interpretive rigor’): 1. Refocusing: “Inconsistency might be a diagnostic tool for detecting pos‐ sible problems in data collection and analysis, or the inferences derived from the results of one strand or the other.” 2. Degree of complementarity: “revealing two different aspects of the same phenomenon” 3. Elaboration or conditionality: “one set of inferences provides the condi‐ tions for the applicability of the other” They conclude that, if these measures do not lead to a plausible explanation or understanding of the divergence, the findings might reveal that “two different but equally plausible realities exist” (2008: 116; see also Lazaraton 2005: 219; Flick 2011). Applied to the example from my preliminary analysis above, the diver‐ gence could be explained or understood in the following ways: a. The sampling caused divergent findings, since interview participants were not the same as the ones in the survey. However, in terms of context and demographics, participants in both strands did not differ. b. The instrument design or nature of the instruments led to different ans‐ wers. Answering questions in a survey can be affected by factors such as fatigue, the time provided to think about the question, or social desirabi‐ lity (see chapter 5). Answering a question in an interview and being asked to elaborate on the answer might require a more critical or in-depth awa‐ reness of e.g. changes in the learning style und thus lead to a different answer than in a survey. c. Another possibility related to the different instruments could be the for‐ mulation of the questions. While the survey questions or statements were more open to interpretation, interview questions generally were aided by prompts and probes (see chapter 6), which provided the participant with more contextual information. 86 4. Methodology, Research Design, and Research Context 2 Here, in line with the definition of the European Commission (2006: 2) “all forms of learning undertaken by adults after having left initial education and training, however far this process may have gone (e.g., including tertiary education).” d. Related to b): Asking the question(s) in the context of different instru‐ ments could have tapped into or revealed different aspects (regarding the depth or awareness) of the (temporal) language learner self-concept. Going through some of the steps with this brief example from my study already reveals the complexity that goes into the integration of both qualitative and quantitative strands in an MMR study. This is particularly the case when, as it was the case here, the findings show inconsistency. The discussion of the example and possible reasons for divergence has also shown that asking What went wrong? may not always be the appropriate reaction to inconsistency: “In‐ consistency, or divergence, may in fact be considered an advantage of mixed methods. One of the major values of mixed methods research might lie in specific instances in which the two sets of inferences do not agree with each other.” (Teddlie & Tashakkori 2008: 115 f.) As we will see in the final interpretation stage of this research project (chapter 7), it is sometimes precisely this “friction” bet‐ ween quantitative and qualitative strands that make MMR findings so unpre‐ dictable, as the quote by Bryman at the beginning of this section states, but is also a chance for expanded understanding of the complexities in the foreign language classroom (Knorr & Schramm 2016: 92). Thus, employing MMR in this research project has the potential to open up new ways of viewing the language learner self-concept of the young-old. 4.3 Research Context: Adult Education Centres in Germany Lifelong learning 2 has grown in importance in recent decades ( Jarvis 2009a; Waxenegger 2011; Aspin et al. 2012; Moody 2010: 485; European Commission 2006 & 2016; Komp & Aartsen 2013). The European Commission first made it its goal in 1995 ( Jarvis 2009b: 271) to increase the number of adults participating in lifelong learning scenarios in order to “acquire new work skills, for active citizenship, or for personal development and fulfilment,” with the latter reasons of ‘personal development and fulfilment’ cited as being of special meaning to young-old learners. In terms of European initiatives that do have a more explicit focus on older learners, a few shall be briefly described here. One example is a strategy for implementing lifelong learning in the UNECE region (including Germany and Austria) started in 2002. Its aim was to address the challenges 87 4.3 Research Context: Adult Education Centres in Germany posed by social and demographic shifts (Waxenegger 2011: 115). In a similar vein, the European University Association developed the European Universities’ Charter on lifelong learning in 2008 to prepare European citizens for future challenges in the realm of globalisation, the demographic shifts, and technolo‐ gical advances through higher education (Waxenegger 2011: 117). To reach this goal, universities agreed to provide better access to higher education for older learners by, for instance, adapting their enrolment procedures and course of‐ ferings to adults (re-)turning to higher education later in life. Initiatives like these show a growing awareness in Europe for fostering and unleashing the potential of its older citizens, who are no longer part of the workforce. Yet, despite the enthusiastic attempt to give more prominence to lifelong learning in Europe in the first decade of the 21 st century, the European commission (2009) acknowledged that only few changes have been fully implemented in educati‐ onal settings and its goals for 2010 were not met. Another deficit of the lifelong learning policies and statistics provided by the European Commission is that they have their major focus on providing more adults aged 25-64 with access to lifelong learning and, with this, more vocati‐ onal-related goals in mind rather than personal development and fulfilment ( Jarvis 2009b: 11, 272; European Commission 2006: 11 f.). What becomes appa‐ rent here is the limited age range that excludes EU citizens/ learners over 64, which goes against the Commission’s claim to approach lifelong learning “from cradle to grave” (European Commission 2009: 7). While this limited focus ho‐ pefully does not imply the end of adulthood or life after 64, it does exclude a high number of lifelong post-retirement learners and thus blurs the actual life‐ long learning situation in Europe, providing another reason for taking a closer look at lifelong learning after entering retirement. Lifelong learning can take the shape of incidental (informal) learning situa‐ tions in everyday life, non-formal learning, and institutional (formal) learning settings. This research project focuses on young-old language learners in a formal lifelong learning context. The German adult education centre Volkshoch‐ schule (vhs centre; lit. trans. people’s high-school) has been chosen as an example for a formal lifelong learning institution. As has been indicated by some young-old learners in this study’s questionnaire (see chapter 5), for many of those language learners - particularly in rural areas - it is the most important and often the only place providing them with opportunities to learn a foreign language. Taking a closer look at the research or learning context Volkshochschule can provide us with answers to the following questions: Under what circumstances do the young-old learn English as a foreign language at this institution? In what 88 4. Methodology, Research Design, and Research Context 3 It was reported in the press that the budget debate of the Bavarian state parliament was reportedly ‘only vhs centre-level’. They had conducted a publicity campaign in the whole city, built up professional PR, assessed their courses with a strict quality controls, and then this. way does this context have an impact on the quality of language teaching? How does it influence the young-old’s language learning process and how they per‐ ceive themselves as foreign language learners? It can also reveal limits or po‐ tential problems for conducting my research in this area. In the case of my re‐ search project, the German adult education institution Volkshochschule has more than 900 centres all over Germany: a better understanding of vhs centres and their learners, is a crucial step towards a clearer picture of adult education and lifelong learning in Germany. In the following, I will move from more general information of the German vhs centre to familiarize the reader with this type of adult education institution - such as its history, its current situation, and the overall course programs - to more language learning-specific issues as, for instance, the role of language learning and English as a foreign language (EFL), the course structure and lang‐ uage learning material, as well as the English teachers and their learners. Along with national statistics regarding vhs courses, teachers, learners etc., I will also include the statistics of the German state of Hesse, since the data for this research study were collected there. 4.3.1 Volkshochschule - The Institution Da stand tatsächlich wieder in der Zeitung, die Haushaltsdebatte im Bayerischen Landtag habe leider “nur Volkshochschulniveau” gehabt. Eine stadtweite Imagekampagne haben sie gemacht, eine professionelle Pressearbeit aufgebaut, ihr Kursprogramm einem strengen Qualitäts-TÜV unterworfen, und dann so was.  3 - Wiarda 2010 - First of all, what exactly is a Volkshochschule (vhs centre)? The main idea of vhs centres is to give all people - regardless of their age or their previous education - access to lifelong learning or further education. What comes closest to German Volkshochschulen originated from Scandinavian countries - Denmark, Sweden, and Norway (Kulich 1964; Greenberg 1991; Bagley & Rust 2009). The University of the Third Age in France also initiated a similar approach to vhs centres in 1975 by offering learning opportunities without barriers (Whitnall 2012: 653; Moody 2010: 485 f.). If in other parts of the world we try to find institutions equivalent to adult education institutions in English-speaking countries, com‐ 89 4.3 Research Context: Adult Education Centres in Germany munity colleges and Elderhostel come close to the system or structure of German Volkshochschulen (Raby 2009; Whitnall 2012; Moody 2010). Because Volkshochschule is a nationwide concept with more than 900 centres (including 16 coordinating vhs centres for each German state) all over Germany and more than six million participants each year (Haustein & Mischke 2011; Huntemann & Reichart 2014; Bastian et al. 2004), these centres are particularly valuable for large-scale assessment the situation of adult education and the life‐ long learning of adults in Germany. Moreover, they are the biggest supplier of further education for older people, with steadily increasing numbers of partici‐ pants aged 50 years and over (Friebe 2007: 18; Schmidt-Hertha 2014: 24). This is also the reason why Kade (2009: 58) referred to vhs centres as schools for the young-old. But even though a degree of uniformity is provided by the fact that the institutional framework is the same on a national scale, differences become apparent on a smaller, regional scale: course programmes, course contents, ma‐ terials, and teachers adapt to the local demand for lifelong learning. Despite the large number of vhs centres and participants, the current situation and image of the centres in Germany are in crisis. In the public mind vhs centres have the image of providing extremely slow-paced and low quality education for bored housewives or retirees (Wiarda 2010). One such display of this asso‐ ciation occurred in 2014 when Helmut Kohl, former German chancellor, tried to insult Wolfgang Thierse, the former president of the German Bundestag, by referring to him as Volkshochschulhirn (lit. transl. folk high-school brain; Wei‐ land & Wittrock 2014). Interestingly, this was followed by outraged reactions on twitter (#volkshochschulhirn) by vhs centre representatives and also from former and current vhs centre learners expressing that having a vhs brain is not an insult at all, but rather a compliment. The question behind this is in what way the perception of vhs centres and their quality may influence the perception of young-old language learners. If the widely held opinion of language classes at these centres is not a good one or one that does not take language learning and teaching at this institution seriously, how can the learners take their lear‐ ning seriously? In other words, we need to ask whether and to what extent does the public image of vhs centres influence the young-old’s language learner self-concept. 4.3.1.1 Course Programme What do vhs centres have to offer for the general public, i.e. its current and potential learners? Its course programme is varied and this is what, according to Wiarda (2010), secures its success: it can adapt to local demands and interests easily. The downside of this, however, is that participants living in more rural 90 4. Methodology, Research Design, and Research Context areas - and this applies to the ageing population more than to younger gene‐ rations (Haustein & Mischke 2011) - can only choose from a limited range of course options in contrast to the multitude of courses offered in bigger cities such as Munich (Wiarda 2010). Nevertheless, what vhs centres in Germany do have in common are the general course programme areas (or departments). The vhs course programme is divided into six different areas of studies: 1. Politics, society, environment (e.g. classes on historical developments in a particular area, environmental projects) 2. Arts (crafts such as pottery, painting, etc.) 3. Health (sports classes such as yoga, dancing, etc.; cooking classes; psy‐ chology classes) 4. Languages (popular languages, such as English, Spanish, French; but also rarely studied languages such as e.g. Danish, Arabic, Persian, Japanese; German as a second language classes for migrants) 5. Job-related skills (computer classes, interview preparation, etc.) 6. Basic education and degrees (a second chance for getting a high-school or other related degrees) According to the statistics for the year 2013, the study area of languages had the most classes to offer (43.5%), followed by health (20.3%) (Huntemann & Reichart 2014; Huntemann 2014). Very few courses have instructors who are trained tea‐ chers (8.6%) or whose day job is teaching at German public schools (3.7% out of that 8.6%). Unfortunately, the statistics do not differentiate this into different areas. It would be interesting to see, for instance, how many trained teachers (i.e. with a university degree or equivalent qualifications to teach foreign lang‐ uages at German public schools) of that 8.6% are working in the foreign lang‐ uages area. In terms of demographics, most learners attending classes at vhs centres are female (75.7%) and especially older people (50 years and older) make up the greater proportion of all participants at vhs centres (41.5%). More than half of the vhs centres in Germany are operated by the local communities (54.2%). Regarding their financial basis, the centres mainly rely on public support (40.8%) and course fees paid by course participants (40.2%). The course fee for 45 minutes of English at a vhs centre in the state of Hesse, is between approx. 2.30€ and 3.60€ (depending on the region or city, the number of participants and type of class/ level, usually excluding material). The times and duration of classes also vary from e.g. 14 weeks of 1x90 minutes per week, 10 weeks with 1x180 minute class per week, and 2-week intensive courses with several hours of class every day (usually during school holidays). For people who rely on public financial support (due to e.g. low income or unemployment), 91 4.3 Research Context: Adult Education Centres in Germany a concession of 33-75% is available - again, depending on the individual vhs centre or region. Keeping this in mind, experts expect a growing number of people to slip into poverty on reaching retirement in the upcoming years, be‐ cause of the increasing instability of employment, lower wages, and the overall demographic development in Germany (Haustein & Mischke, 2011: 63). Thus, in the future, the attendance of young-old learners in a vhs class will increasingly rely on whether they can afford it - with or without concession. Thus, we can expect a shift of future young-old generations (depending on their digital lite‐ racy levels) towards a preference for cheaper digital language learning instead of attending vhs classes (see chapter 8). Furthermore, the factor of decreasing mobility - especially if a young-old learner is from a rural area and getting to the vhs centre venue is only possible by car - might add to this development. The only major factor which will keep current and future young-old learners from relying much more or only, on the Internet, is the social aspect of attending classes (Murray 2011). Another interesting aspect of the course programmes at vhs centres indicated by these statistics is the large number of classes directed at particular target groups. The salient target groups in the statistics are classes for migrants and classes for older people (Huntemann 2014). However, as noted by Friebe (2007), fewer older people are tending to take these aged-specific classes over the years. It is a trend which has also been confirmed by programme coordinators inter‐ viewed for my diploma thesis (Neigert 2011). In the 1990s and at the beginning of the 2000s many course titles included the term ‘seniors’, e.g. English for se‐ niors. Because of the negative connotation of the term ‘senior’, however, fewer aged people were attracted to this type of course. ‘Senior’ was soon replaced in many vhs course programmes with a more “neutral” term using the chronolo‐ gical age (e.g. English 50plus). But even this new name caused some confusion among potential participants: “If I am 48, can I still join the 50plus class? ” One of the programme coordinators in my study in 2011 said that, rather than indi‐ cating the target group in the course title, they are trying to move to addressing older participants by the allocated course times. According to him, classes in the morning or throughout the day are generally more appealing to older people than classes taking place in the evening. After taking a more general look at the different subjects a vhs centre has to offer, the next step is to look closely at the subject foreign languages and how they are taught at vhs centres. In order to do so, the next section will explore this study area by taking a closer look at important factors such as contents (topics and materials), the teachers, and the learners of English as a foreign language. 92 4. Methodology, Research Design, and Research Context 4 See: In Conversation (Klett) or Network B1 Refresher (Langenscheidt) 4.3.1.2 Learning Foreign Languages at vhs Centre: Contents, Teachers, and Learners There is certainly no particular point in the ageing of healthy adults where L2 learning becomes impossible. - Singleton & Ryan 2004: 214 - As mentioned above, the foreign language sector of all study areas at vhs centres offers the largest number of classes (43.5% of all classes offered) and usually has the second-highest number of enrolments (after health-related courses). What characterizes learning foreign languages at vhs centres compared to other lang‐ uage learning institutions for adults or learning languages at a public school in Germany? If we take a look at the statistics again, English, followed by French, Italian, and Spanish, is the most popular language. Of all three latter languages, ho‐ wever, only Spanish has seen a steady increase in courses offered between 1987 and 2008 (Ambos et al. 2010: 5). While originally structuring the classes by using the terms beginners, intermediate, and advanced, more and more vhs centres have adopted the categories provided by the Common European Framework (CEF levels: beginner: A1, A2; intermediate: B1, B2; advanced: C1). In addition to this, they used to offer more targeted group-specific titles which were, as mentioned above, often the result of addressing or adapting to local demands (e.g. conversation classes, business English, English for travel, preparation classes for tests or certificates such as TOEFL or IELTS, English for seniors). The target group-specific options still exist, but in more recent years they are offered in combination with CEF levels (e.g. A1 English 60plus, B2 English conversation group, B1 Business English Refresher). Again, learners in more rural areas have a limited access to furthering their language proficiency beyond certain levels as there are simply not always enough interested learners in the area to reach the minimum number of parti‐ cipants needed for starting a class (usually at least 6 or 8 participants). Textbook publishers, it seems, have adapted their product range to these course offers such as textbooks targeting adult language learners in conversation courses, refresher courses, and business English courses. 4 It is also possible, however, that this is a mutual process: on the one hand, there are vhs centres offering classes based on existing textbook formats (e.g. a course called Sterling Silver at a Hessian vhs centre named after the respective textbook for older learners be‐ ginning to learn English) or certificates (e.g. TOEFL or IELTS) and, on the other 93 4.3 Research Context: Adult Education Centres in Germany hand, there are textbook publishers trying to address the needs and interests of local vhs centres and their customers. From these initial descriptions of (English) language class characteristics at vhs centres, some differences to foreign language classes offered at German public schools become apparent. Firstly, the learner - or, in the case of vhs centres, customer - has more impact on the courses offered at vhs centres than students at high school. At German public schools, more general decisions on the curriculum are made at a state level. Specific decisions regarding which textbooks are to be used are made at the (language) teacher conference level (often textbooks from a certain publisher will be used for years or decades). This is often done in order to achieve uniformity within schools. Decisions on course format and contents are more of a top-down process. At vhs centres, this process is somewhat reversed since the existence of classes is dependent on participants and the money they pay for the course (as mentioned above: 40.2%). Although the teacher and learners in each class have a considerable amount of influence on the contents and material used, most vhs centres are trying to attain a more uniform structure and contents - especially at the beginner level - within in‐ dividual institutions in terms of such things as textbook choice. As already addressed in the previous section, another major difference bet‐ ween foreign language classes at German public schools and vhs centres are their instructors or teachers. Merely 8.6% of all vhs instructors in Germany are trained full-time teachers (this includes teachers from all study areas, not only foreign languages). This may come as no surprise, as teaching classes at a vhs centre is usually not suitable for a full-time job, but rather as a side-job, free‐ lancing or even working voluntarily (Bastian et al. 2004). The average vhs centre instructor earns 18€ (25€ for German as a foreign language/ integration classes) for 45 minutes (Wiarda 2010). The majority of vhs teachers in Germany work on a casual or voluntary basis (188,058 teachers in 2013), whereas only 3,289.2 teachers work full-time (75.3% of which have permanent positions; Huntemann & Reichart 2014). The backgrounds of teachers and reasons for teaching at a vhs centre are thus multifarious. Some vhs centres prefer, for example, native spea‐ kers as teachers for foreign language classes, often regardless of their qualifi‐ cation. Some other vhs centres in the state of Hesse make it compulsory for new language teachers to attend a workshop introducing them to the basics of foreign language teaching in adult education. Fairly recently, vhs centres all over Ger‐ many offer their course instructors a further education or qualification pro‐ gramme called Erwachsenenpädagogische Qualifikation (EPQ; transl. lit.: adult pedagogical qualification; Pfirrmann 2015) consisting of different modules such as an introduction to adult pedagogy (or: andragogy), subject-specific works‐ 94 4. Methodology, Research Design, and Research Context hops, a report reflecting on the teaching experiences, and a colloquium. Even though this qualification programme is not obligatory for teachers, it is an at‐ tempt to work towards a homogenization of teaching performance and to foster teaching quality. The existing heterogeneity of teachers and course contents enables vhs centres to adapt to local needs. One could even argue that teachers with different kinds of backgrounds outside university initiate, or are more open to, innovative teaching concepts. The downside of this, however, is that getting good quality or professional language instruction - especially in rural areas where more young-old people tend to live - is a matter of luck. Providing vhs teachers with opportunities for further education or training can help reduce this risk. Mo‐ reover, these opportunities to further develop teaching skills and vhs teachers’ knowledge of adult education and more specifically language teaching/ learning, often give them the only space where they can exchange their language teaching experiences and ideas with colleagues. In contrast to teachers at public schools, teachers at vhs centres rarely have faculty rooms to meet and to prepare classes. Thus, workshops for teachers at vhs centres, which are also an organized way of networking, are valuable for their development, both professionally and so‐ cially. Expanding one’s skills set for language teaching and feeling less isolated in this job can instil greater confidence among vhs teachers and thus help im‐ prove and sustain their language teaching quality. Even though the increased teacher training possibilities promise improve‐ ment, vhs centres in Germany experience greater fluctuation among teaching staff than public schools. This may partly be due to the different backgrounds of teachers and their working conditions. As mentioned above, teachers at vhs centres mainly work as casuals on the side and, additionally, in rural areas with few participants, there is less stability for many classes, which means less sta‐ bility of income for teachers. But despite that, some classes or learner groups grow together and remain together for decades. Several participants of two En‐ glish conversation classes interviewed for this study have been part of their class since the 1980s and 1990s. They have seen many other learners and different teachers come and go. Furthermore, in contrast to public schools, despite the stronger tendency to fluctuation of teachers (and often learners), a core of young-old learners often stay in their English class. The analysis of their inter‐ views (see chapter 6) will shed further light on this phenomenon. What about the learners (customers)? The term “customer” applies here be‐ cause, after all, in contrast to public schools, language learners pay for the service of being taught a language (Bastian et al. 2004). Referring to statistics again, the study area of foreign languages at vhs centres has more female (67.9%) partici‐ 95 4.3 Research Context: Adult Education Centres in Germany pants than male. As mentioned above, vhs classes in Germany generally have more female (75.7%) than male participants (this applies to the sum of all parti‐ cipants regardless of age). In terms of age, 22.8% (Hesse: 21.0%) of language class participants are between 50 and 64, and 14.4% (Hesse: 13.1%) are 65 years and older. Combined, language learners in the 50plus group make up more than one third of all language learners at vhs centres in Germany, and this number has been increasing (Völkening 2006; Eschmann 2005). Does this have noticeable effects on the foreign language area at vhs centres? The results of a small-scale survey for my diploma thesis indicated surpri‐ singly that the main reason given for attending EFL classes at vhs centre, by young-old participants was keeping their mind active (followed by travelling as another major reason for learning English), which is considered by some young-old language learners to be an effective method of dementia prevention. The earlier survey (Neigert 2011), but also the present study, showed that very few young-old learners were interested in learning English or receiving prepa‐ ration for tests for English certificates. Most young-olds were no longer working and could finally spend their free time in the way they wanted and approach language learning with less pressure. Vhs centres have responded to this wish or preference by including this in some course descriptions for their 50plus or 60plus, by classes including such key words as “relaxed”, “slow paced”, or “plea‐ sant atmosphere.” As mentioned above, the young-old make up more than one third of language learners at vhs centres, which rely heavily on their fees; the existence of courses sometimes depends on their staying due to the aforemen‐ tioned minimum number of participants required to continue offering a course, making it particularly difficult to offer and maintain certain language courses. It follows from this that the young-old possibly have a greater impact on the language class design in terms of topics, material, and pace. What could the positive and negative consequences of this be? 4.4 Summary and Outlook Studying and understanding young-old language learners’ self-concepts, and in particular their temporal aspects, can help us predict what they expect from themselves, from the institution of vhs, and from their teacher in the realm of language learning. Understanding their self-concepts in the domain of foreign language learning in view of their past, present, and future possible selves can shed light on what keeps them in a language class, but also on which factors of their possible selves probably discourage them. Reaching this level of under‐ 96 4. Methodology, Research Design, and Research Context standing requires measures that can capture the complexities of the language learner self on both a largeand small-scale level. Therefore, this chapter out‐ lined why and how using mixed methods provides a means of working towards a better understanding of young-old language learners. Both quantitative and qualitative strands offer different perspectives on the quite complex issue of language learner self of young-old learners. The survey (QUAN strand) in the next chapter, for instance, offers a broad look at L2 self-concept based on a larger sample of learners (n = 195) at several vhs centres in Germany. The interviews (QUAL strand) in chapter 6 provide an in-depth exploration of 21 young-old learners’ L2 self-concepts in three different classes at one vhs centre in Germany. Most importantly, however, this chapter has il‐ lustrated that the combination of the two strands (in chapter 7) extends indivi‐ dual perspectives on the L2-learner self in terms of how and why they converge and diverge. 97 4.4 Summary and Outlook 5. Numbers: The Survey As I have outlined in my chapter on methodology (chapter 4), the survey re‐ presents the quantitative (QUAN) strand of my mixed methods design. This chapter addresses the processes involved in creating the questionnaire, testing it, and sending it out to participants. The development and design section in this chapter, will show how it is often the underestimated small things such as as‐ pects of formulation or layout in a questionnaire can often have a great impact on the results. This chapter also covers the processing and analysis of the com‐ pleted questionnaires using the statistical software SPSS. It will thus begin with the construction of the questionnaire, the administration of the survey, and the eventual processing of the completed questionnaires. A final section in this chapter presents a detailed analysis of the survey results. What makes questionnaires valuable for my research project and what are the potential pitfalls in employing them? Questionnaires are highly structured instru‐ ments (Dörnyei 2010: 9): they are the main or only interface between researcher and respondents, meaning that their design and the formulation of questions need to be considered carefully in order to avoid misinterpretation among respon‐ dents. Fortunately, it is this high degree of structure that also makes question‐ naires compatible with quantitative research. According to Dörnyei (2010: 9), quantitative research “employs categories, viewpoints, and models that have been precisely defined by the researcher in advance, and numerical or directly quanti‐ fiable data are collected to determine the relationship between these categories and to test research hypotheses.” Thus, questionnaires are a versatile and effective in‐ strument to test research hypotheses and to find correlates of various factors for a relatively large group of people (Cohen et al. 2011; Dörnyei 2010: 6; Gillham 2000: 6). In this way, they provide a good “counterbalance” for the qualitative strand of my research project, i.e. interviews with 21 young-old learners. Furthermore, Nunan and Bailey (2009: 125) describe the purpose of questionnaires as a “snapshot of conditions, attitudes, and/ or events” by eliciting these from a sample of a popu‐ lation. If the data collection and analysis of questionnaire data are carried out ap‐ propriately, i.e. if validity criteria are met, the results can be applied to larger po‐ pulations. In the case of my research project, survey participants constitute a sample of young-old language learners of English as a foreign language (EFL) at different vhs centres in the German state of Hesse (see also the sub-section on “Research Context” in chapter 4). But a questionnaire’s effectiveness in terms of time, effort, and money comes at a price. Questions and their possible answers in a structured instrument such as a questionnaire (Gillham 2000: 2 f.) are predetermined by the researcher and thus, [T]he element of discovery is much reduced (unless there is a very unexpected pattern to the answers selected). You don’t know what lies behind the responses selected or, more importantly, answers the respondents might have given had they been free to respond as they wished. In other words, as already discussed in the previous chapter, the results of qu‐ estionnaires may derive from and be applied to a broad spectrum within a certain population. But at the same time, they lack depth in contrast to interviews. The results are limited to, or hinge on, the predetermined questions and answers, which are potentially contextualized differently by each respondent. As we will see by the end of this chapter, the scope of a questionnaire ends where the insight of interviews begins. As already mentioned in the previous chapter on mixed methods, the research questions need to be reformulated into hypotheses, which are a more appro‐ priate form of research questions in QUAN-studies. Related to this, the second sub-question of my research questions (see also section 5.1 below) strongly points to the aims of this quantitative data strand: Which factors (e.g. age-related: retirement) influence the young-old learners’ self-con‐ cept? Correlating demographic data of survey participants with their responses to the questionnaire items, therefore, can help find answers to this question as well as to the overarching research question on gaining a better understanding of the self-concept of young-old language learners. Thus, the hypotheses formulated below explicate the influential “factors” in the above research question and are based on the theoretical implications in earlier chapters of this thesis. As Ryff ’s study on possible selves (see chapter 3.3.1) and age spans has indi‐ cated, there has been a higher past possible self in older participants. Further‐ more, the possible self scores between male and female participants have dif‐ fered significantly. In addition to this, the mathematical and verbal self-concept scores of male and female students in Marsh’s study have differed as well, with the latter showing a preference or higher scores for verbal self-concept (see chapter 3.2.3). Thus, both studies indicate a requirement that a hypothesis for the purposes of my study tests possible female/ male differences of young-old language learners. A hypothesis which also explores the sociological dimension 100 5. Numbers: The Survey of age and its potential impact on young-old learners’ self-concepts is also needed. Chapters 2 and 3 have both highlighted the relevance of the sociological age marker and major life transition “retirement”. Another relevant factor that may influence self-concept and that demarcates young-old learners from yo‐ unger learners at school is the relatively long time some of them spend learning English at the vhs centre. As mentioned in chapter 4.3.1.2, some of the young-old learners interviewed for this study as well as some in the survey study (as the descriptive statistics below will show) have spent one or more decades in their English class. It may therefore be interesting to compare their responses with the ones of young-old learners who have only recently joined their vhs centre English class. From the research questions in the methodology chapter and the theory presented in chapters 2 and 3, I derived the following hypotheses for the quantitative data analysis purposes: 1. Past and present L2-self scores are higher than the future/ ideal L2 self among young-old learners. This discrepancy increases with age (i.e. when comparing different age groups among the young-old language learners). 2. Being retired has a negative impact on L2 self-concept and its temporal aspects. 3. The time spent in a vhs centre English course has a positive effect on L2 self-concept. 4. Female language learners have higher L2 self-concept scores than their male counterparts. It is the aim of the data analysis section 5.4 to find answers to these hypotheses and, in this way, to the main research question about young-old language lear‐ ners’ self-concept. Before going into the results of the survey, however, the next sub-chapters first give an insight into how the questionnaire items were deve‐ loped based on the theory and research aims, and how the questionnaire was designed and administered. 5.1 From Theory to Instrument: Development of Questionnaire Items In the development phase of my questionnaire, I turned to already existing survey studies for the collection of appropriate items, which would help me “operationalize” (Nunan & Bailey 2009: 130) the L2 self-relevant constructs addressed in my research questions: 101 5.1 From Theory to Instrument: Development of Questionnaire Items How do the young-old conceptualize themselves as foreign language learners? a. How do they perceive their past, present, and future L2-self/ selves? b. Which factors (e.g. age-related: retirement) influence their language learner self-concept? c. In what ways do survey and interview study converge? As pointed out in chapter 3, however, existing questionnaires on language le‐ arner self-concept are predominantly directed at children and young adults. If adult learners were addressed, the age range of the participants rarely went beyond 30 or 40 (Byrne 1996a; Weil 2015; Mercer 2011a; Marsh & Ayotte 2003; Dörnyei 2010). Thus, parts of the already existing item pools needed to be adapted to the perspective of my study participants. For example, in the item pools developed by Taguchi (Dörnyei 2010), participants are asked about the role of parents in the language learning process at university (e.g. expectations, pressure, decision-making, motivation - in other words, mainly aspects of the realm of ought-to selves). In view of the young-old participants in my study, their parents do not play a role as a socializer since they are either usually the ones dependent on their young-old children or are no longer alive. Thus, I have included the factor of grandchildren and partners in several questions in the questionnaire. In the questionnaire’s development phase, I collected around 50 different items for the section using the Likert scale. The items were translated into German, adapted from Taguchi’s contributions to Dörnyei’s book (2010: appen‐ dices A & B), and some were based on the interview questions in Mercer’s study (2011a: appendix C). Since the latter consisted of open-ended interview ques‐ tions, the questions relevant for my study, i.e. for answering my research ques‐ tions above, needed to be reformulated as statements suitable for a Likert scale. For example, the question in her interview guide “Do you think you’re a ‘typical’ language learner? Explain.” became “I consider myself a typical language le‐ arner.” (question 22) in my questionnaire. This item conceptualized the general and actual/ present facets of the young-old language learners’ selves addressed in research question a. above. Drawing on items from existing surveys necessitated following the advice of Sudman and Bradburn (1983: 119, as cited in Dörnyei 2010: 24): The best advice we can offer to those starting out to write attitude questions is to plagiarize. While plagiarism is regarded as a vice in most matters, it is a virtue in questionnaire writing - assuming of course, that you plagiarize good quality ques‐ tions. 102 5. Numbers: The Survey Further, drawing on questions from existing studies also meant that they had been tested for their purposes and, to some extent, made it possible to compare study results afterwards. The Likert scale items were grouped around different concepts based on tem‐ poral aspects of self-concept as identified in chapters 2 and 3: general self-con‐ cept, past-, present-, and ideal-L2-selves, perceived learning progress and impact on cognitive abilities, as well as learning environment. General self-concept items addressed self-concept aspects of young-old learners that were not directly related to learning a foreign language but to learning in general, as well as to the way young-old respondents perceive themselves (e.g. in terms of self-con‐ fidence) in general. Past, actual, and ideal L2 self items were based on the theories of possible selves and - in relation to learning a foreign language - Dörnyei’s future self-guides (ideal and ought-to L2 self) within his L2 Motivational Self-System (L2MSS). The operationalized L2 construct designated as the “im‐ pact on cognitive abilities” is derived from the active ageing trend discussed in chapter 2 and it also ties in with Dörnyei’s ought-to L2 self presented in chapter 3. Items under this heading capture the young-old learners’ intention to learn a foreign language to keep their minds active. As the descriptive statistics below indicate, this motive seems to be particularly strong within this target group (see section 5.4). Finally, questionnaire items for the operationalized const‐ ructs “perceived language learning progress” and “learning environment” are derived from the L2MSS component of “language learning experience”. Accor‐ ding to Dörnyei, this component encompasses other language learning aspects such as the learning context, teacher, experiences of success using the L2 etc. table 5.1 “Likert Scale Items and Operationalized L2-Constructs” below shows a complete list of the 28 Likert items in the order they are used in the final version of the questionnaire and it indicates which operationalized L2-self constructs are addressed by each item. As an example, figure 5.1 below illustrates how the concept of “ideal L2 self ”, introduced in chapter 3, has been operationalized as Likert item number 16 in the questionnaire and is represented in table 5.1 below: Figure 5.1 Sample Item for Ideal L2 Self 103 5.1 From Theory to Instrument: Development of Questionnaire Items 1 Translated and adapted from Dörnyei and Taguchi’s Ideal L2 Self item list (2010: 140). For example, “living abroad” was changed to “when travelling abroad” as this seemed closer to the young-old participants’ everyday lives. In this way, together with items 10, 13, and 24 in the questionnaire, a group is then formed, which addresses aspects of the ideal L2 self of the survey partici‐ pants: 10. I can imagine a situation where I am using English with people from abroad. 13. There will be situations that require me to use English outside the classroom. 16. I can imagine myself using English when travelling abroad and to effectively communicate with the locals. 24. I can imagine myself speaking English like a native speaker.  1 As the different numbers of the item group for ideal L2-self example reveal, mixing the questions of different item groups is another strategy to mitigate leading participants significantly in their answers. If, for example, all questions of one group were listed sequentially in the questionnaire, participants would become aware of the internal structure and would thus put less effort into ans‐ wering the individual questions more carefully. The strategy of grouping items targeting abstract phenomena such as self-concept is referred to as multi-item scales. Since a single question item can be interpreted differently - even if it is carefully formulated - multi-item scales use the sum of a group of items targe‐ ting that particular phenomenon and in this way “any idiosyncratic interpreta‐ tion of an item will be ironed out during the summation of the item scores.” (Dörnyei 2010: 24). Nevertheless, this is not to say that responses to individual Likert scale items cannot also already reveal interesting information about the temporal facets of the language learner self-concept. For this reason, the final merging of quantitative and qualitative data strands (see chapter 7: micro-level results) will also include the comparison/ contrasting of single Likert scale items with individual qualitative themes. Table 5.1 on the following page shows all 28 Likert scale items in the order they appear in the questionnaire and the L2-const‐ ructs they operationalize. 104 5. Numbers: The Survey Items (translated from German) Operationalized L2-Constructs General selfconcept Past L2 Self Actual L2 Self Ideal L2 Self Lear‐ ning pro‐ gress Impact on cog‐ nitive abilities L2 Lear‐ ning en‐ viron‐ ment I’m looking forward to my English class. x x I enjoyed going to school. x x It is no problem for me to use English in class. x I feel that I’m making progress with my English. x x I enjoy speaking in class. x x I’ve always had a talent for languages. x I’m happy with my learning strategies. x I’m ambitious when it comes to studying Eng‐ lish. x x It is hard for me to study English. x x I can imagine a situation where I am using En‐ glish with people from abroad. x My English class is what I expect to be an ideal foreign language class. x I’m working intensively on my English skills. x x 105 5.1 From Theory to Instrument: Development of Questionnaire Items Items (translated from German) Operationalized L2-Constructs General selfconcept Past L2 Self Actual L2 Self Ideal L2 Self Lear‐ ning pro‐ gress Impact on cog‐ nitive abilities L2 Lear‐ ning en‐ viron‐ ment There will be situations that require me to use English outside the classroom. x I think I can speak English well. x x Whenever I encounter problems while studying English, I give up quickly. x x I can imagine myself using English when tra‐ velling abroad and effectively communicating with the locals. x I approach learning English in a more relaxed way than in the past / at school. x x I feel mentally more active and fitter because of studying English. x x The vhs centre has the best learning conditions. I’m a self-confident person. x x Learning English comes naturally to me. x x I consider myself a typical language learner. x x Learning English helps me to feel younger and fitter. x x 106 5. Numbers: The Survey Items (translated from German) Operationalized L2-Constructs General selfconcept Past L2 Self Actual L2 Self Ideal L2 Self Lear‐ ning pro‐ gress Impact on cog‐ nitive abilities L2 Lear‐ ning en‐ viron‐ ment I can imagine myself speaking English like a native speaker. x I’m happy with my current language level. x x The way English is taught at the vhs centre is comparable to the English classes I had at school. x There are no comparable language learning op‐ tions other than the vhs centre for me. x The way I study English has changed over time. x Table 5.1 Likert Scale Items and Operationalized L2-Constructs 107 5.1 From Theory to Instrument: Development of Questionnaire Items 5.2 The Questionnaire Design Because in surveys employing self-completed questionnaires the main interface between the researcher and the respondent is the hard copy of the questionnaire; the format and graphic layout carry a special significance and have an important impact on the responses. - Dörnyei 2010: 13 - Regarding its structure, I have included different question formats in the ques‐ tionnaire (see appendix A) and divided it into several parts, framed by a greeting, instructions and a final thank you to the participants (Reinders 2011a). The greeting plays an important part in a questionnaire, because it sets the stage for what is to come and gives background information to respondents. For example, it tells the participant who is involved in the conduct of the survey the reasons and goals in conducting it, e.g. in this case the improvement of learning condi‐ tions for adult language learners at vhs centres. In this way, participants know for what purposes they are “sacrificing” their time. When formulating the gree‐ ting or introduction to my questionnaire, I took great care to use comprehensible terminology but at the same time not to confuse participants or to lead them too much in a certain direction when later answering the questions (Nunan & Bailey 2009: 126). The greeting also gave an overview of the overall structure of the questionnaire (three parts) and gave participants an idea of how long it would take them to complete it (15 minutes). Participants were also reminded of the importance of their answers being truthful and assured that the questionnaire was not a test, there being no correct or incorrect answers. Finally, participants were also assured of the confidentiality their answers and data and that the survey would only be used within the framework of the dissertational project. The greeting ended with thanking them for their participation and support, which at the same time indicated the start of the three parts of the questionnaire. Before describing each of the three question parts in the following, a few words are needed on the instructions regarding questionnaire items. The in‐ structions were particularly important for participants unfamiliar with comple‐ ting research questionnaires. They needed to be formulated concisely and clearly, giving participants a quick but comprehensible idea of how and why to answer questions (Reinders 2011a). Ambiguous or unclear instructions lead to incorrectly answered questions making the results unusable for further analysis. For instance, part I (Likert scales) and question 3 in Part II (adjective scales) were preceded by an example to illustrate the values of the scale parts (i.e. what does “6” on the Likert scale from 1 to 6 represent? ) and the instructions also 108 5. Numbers: The Survey 2 Translation: “The teaching and learning concept of my English class at the vhs centre corresponds with/ is comparable to the one I experienced at school.” indicated whether participants were required to tick or to circle an answer or a number: Figure 5.2 Likert Scale example illustrating the values of 1 to 6 with the sample sentence "I like to drink tea" (Ich trinke gerne Tee). Here, the pilot study already provided useful feedback on how to reformulate instructions to make them clearer and change the design to avoid confusion. For example, the pilot study showed that the values that the Likert scale numbers represent were only listed on the first page as a reference. I included the refe‐ rence on top of the second page again in order to decrease the unnecessary effort of turning back and forth between pages for checking the values of 1 (does not apply at all) to 6 (totally applies). This is only one of several examples of rela‐ tively small instructional and layout changes that have a big impact: in this case, it makes navigation through the questionnaire easier and frees up more cogni‐ tive capacity so participants can better focus on the questions (Dörnyei 2010). As mentioned, the first question part contained Likert scale items with the scale ranging from 1 (does not apply at all) to 6 (totally applies). However, the use of scaled responses has disadvantages, particularly if items are not carefully formulated. Gillham, for example, points out that participants often do not use the full scale, they tend to be rather positive (e.g. in customer satisfaction sur‐ veys), and - as mentioned above - “whatever response they tick, you don’t know why.” (2000: 32). Using either an odd or an even number for the scale is also a widely debated issue in questionnaire development research (Gilham 2000: 32). By choosing an even number for the scale (1 to 6), I am forcing participants to make a decision in a certain direction. Of course, this does not apply to all qu‐ estion items, which neither the pilot study nor the feedback indicated. Question 26 in the final draft of the questionnaire “Das Lehrkonzept meines Englischkurses an der vhs entspricht dem Fremdsprachenunterricht, den ich in meiner Schulzeit hatte.”  2 assumes that all participants had the opportunity to learn English at school in Germany during and after World War II. However, unlike today, where 109 5.2 The Questionnaire Design 3 Translation: “It is hard for me to study English.” 4 Translation: “Whenever I encounter problems studying English, I give up quickly.” English as a Foreign Language is an indispensable part of the national school curriculum, it was not as common to learn English at school back then. In this case, participants had no opportunity to “opt out” of answering the question by ticking e.g. “not applicable” in the centre. Carefully formulating, combining, and ordering the question items is key to getting the best results. This is why, for example, several more interesting or captivating questions, i.e. questions participants could more easily relate to (such as “I look forward to going to my English class.”; question 1), were included on the first page of the questionnaire (Dörnyei 2010: 47; Reinders 2011a). Ne‐ gative and positive formulation of one and the same concept increases its vali‐ dity. Examples of a negatively formulated question in the item pool are questions 9 (“Englisch zu lernen ist anstrengend für mich.” 3 ) and 15 (“Wenn Probleme beim Englischlernen auftauchen, gebe ich schnell auf.” 4 ). The inclusion of negatively formulated questions like these tackles problems of Likert scale items such as fatigue and acquiescence bias (Dörnyei 2010: 9 & 43; Mummendey & Grau 2008). Further considerations regarding the formulation of items for part I included avoiding negations as well as ambiguous formulations (e.g. two questions or concepts in one). The 50 Likert scale items collected and (re-)formulated in the initial develop‐ ment process needed to be reduced in order to keep the amount of time it would take to complete the questionnaire below approximately 15 minutes, but also for layout-reasons as will be discussed below. Additionally, a reduction of the Likert items provided space for other question formats (e.g. open-ended ques‐ tions or questions on respondents’ demographics). Feedback from fellow-re‐ searchers and the pilot study revealed items that could be regarded as not rele‐ vant to the study’s aims or easily misinterpreted. For instance, one item stating that the participant could identify with the teaching-concept at vhs centres as a foreign language learner led to confusion amongst pilot study participants and was thus removed from the questionnaire. Items about the influence of life ex‐ periences (from work or home) were also removed from the questionnaire since they were ambiguous to many participants, or not regarded as being directly related to the research questions. The pilot study played a pivotal role in the development and further refine‐ ment of the questionnaire. A group of young-old English learners I used to teach as a substitute teacher for one semester volunteered to pilot-test my questionn‐ aire, making this a so-called convenience sampling (Cohen et al. 2011: 155; Gay 110 5. Numbers: The Survey et al. 2012: 141 f.). I was also able to find four volunteers to pilot test my interview questions at a later stage from that same group (see chapter 6). I asked the pilot study participants to highlight any confusing or difficult items or questions as they completed the questionnaire. The follow-up discussion after they had com‐ pleted the questionnaire gave me a good impression of which questions needed to be removed or reformulated. In order to check the envisioned 15-minute time limit for completing the questionnaire, I decided to be present while pilot study participants completed it. Additionally, pilot study participants could use the opportunity to ask me questions while they were completing the questionnaire, which in turn gave me an immediate feedback on the design and contents. The pilot study showed that the original aim of 15 minutes was a good estimate. I did not want to go consi‐ derably beyond that amount of time, as I expected the questionnaires to be completed during class-time. Since the average English class at vhs centres is only 90 minutes a week, time is very precious in this context and I consider devoting 15 minutes to completing my questionnaire a big sacrifice on the side of both teachers and learners. This meant reducing the questionnaire contents to the most relevant constituents. The second part of the questionnaire consisted of three different question formats. The questions in this part were intended to elicit reasons or motifs for learning English at the vhs centre, how young-old participants perceive them‐ selves as language learners in their current vhs centre class and - if applicable - at earlier stages of their lives (e.g. at school). Questions 2 and 3 in the second part especially targeted at the participants’ past and current L2 selves. I decided to proceed from an open question format in question 2 (naming three charac‐ teristics or traits they possess which help them to learn English) to a more closed format in question 3. In question 3 they were asked to rate on adjective scales (see figure 5.3 below; they are also referred to as semantic differential scales, Dörnyei 2010: 29), how they perceive themselves in their English class at vhs centres currently and in past English learning contexts (e.g. at school). At the ends of each scale were characteristics or traits such as active/ passive, coope‐ rative/ independent, introverted/ extroverted etc. Ordering these two questions from open to closed, I hoped that their answers to question 2 would not be influenced by the characteristics/ traits provided in number 3. However, in con‐ trast to online surveys, a researcher cannot control the sequence in which le‐ arners answer the questions in a self-completed questionnaire on paper (Gillham 2000: 12). Unlike the Likert scale in the first part of the questionnaire, the scale in question 3 in the second part consisted of an odd number, thus not forcing an 111 5.2 The Questionnaire Design answer or providing an “undecided” option. Through pilot testing, however, the range of the scale was reduced from seven to five for improved clarity of options. The adjective scales (Question 3) in the second part of the questionnaire posed some problems before and after having the questionnaires completed (see figure 5.3 below for the first adjective scale in question 3). Firstly, in the development phase, finding adjectives for the opposing ends of the scales was difficult. Ad‐ jectives with opposite meaning are often connoted either positively or negati‐ vely, and this way could be subject to social desirability or prestige bias (Dörnyei 2010: 8; Mummendey & Grau 2008) when participants are making their tick on the scale. While some of the chosen adjective pairs are relatively neutral in their meaning (cooperative/ independent; content/ ambitious), others carry a stronger negative and positive meaning (active/ passive, distracted/ focused). Secondly, question 3 was divided into two parts intended to elicit self-perceptions about their current and their past L2-selves. The first part asked participants to indicate how they currently feel about themselves in the vhs class. The second part of question 3, however, turned out to be ambiguously formulated since it asked participants how they feel in their vhs class compared to the time they were learning English at school. The question was intended to elicit information about their past L2-self, but, in fact was asking again about their current self having their past experiences in mind. Feedback and pilot study, however, did not point out that flaw. The ambiguity of the second part of question 3 probably led to participants interpreting the question in different ways, rendering its results unusable for further analysis. Figure 5.3 Adjective Scale in Question 3a The third and final part asked participants for demographic information and again assured participants that their answers and data are handled with care 112 5. Numbers: The Survey and only used for the purposes of the study. Demographic information such as age, gender, time spent in vhs centre English class (providing time frames, if participants are retired or still working, their degree, and profession), was col‐ lected. Dörnyei (2010: 47 f.) suggests leaving this part to the end of the questi‐ onnaire so that, firstly, participants are not “put off ” by being asked personal information right at the beginning, and, secondly, there is space for the more captivating questions at the beginning. Nevertheless, the demographic infor‐ mation helps to make sense of responses: How do the collected answers correlate with a certain age range? Are there differences in the responses given by men and women? Regarding the formulation of questions about their age it showed that only a few participants did not provide their age (usually female). This poses the question of whether participants would feel more or less reluctant to give information about their age if I had asked them for their year of birth instead of their age. The questionnaire was rounded off by a final ‘thank you’ to the par‐ ticipants. The final design and format of the questionnaire was the product of a deve‐ lopmental process (illustrated above) to which feedback from fellow researchers and the pilot study contributed significantly. The questionnaire consists of three central parts, as well as an opening (greeting and instruction) and closing (final ‘thank you’) section. Again, and as Dörnyei in the above quote rightly points out, the layout of a self-administered questionnaire needs to be consi‐ dered carefully, as bad decisions in this regard can have a negative impact on the results. The right layout decisions make a questionnaire look more profes‐ sional, and make it easier for participants to complete the questionnaire; this potentially increases the response rate (DAA 2016). In terms of its length, four pages and an estimated 15 minutes seemed the most appropriate for this target group and context. A length of four pages also made it possible to present the questionnaire in booklet form using one DIN A3-sized paper rather than two DIN A4-sized papers which are printed on both sides and stapled. This gives a more compact impression, it is easier to turn pages, and prevents the potential loss of pages (Dörnyei 2010: 13). Connected to this is the issue of the paper itself. Newell (2012: 109; cited in Dörnyei 2010: 15) argues that even such fine and underestimated aspects of the questionnaire such as the quality of the paper used have an impact on the response rate, which is why she has been using thick, beige paper since it stands out, is pleasant to handle, and “people will not have the heart to throw away such an attractive document.” In line with the last argument, I would also like to argue that using a questi‐ onnaire on paper also tends to be more appreciated than providing a link to an 113 5.2 The Questionnaire Design online survey. Using a paper questionnaire becomes even more valuable when used with this particular target group. While we can assume that all young-old language learners attending an English class know how to handle pen and paper, we cannot expect an equally high digital literacy. Using an online-format for the questionnaire would exclude those English classes or young-old participants without computer access, thus leading to sampling issues. As pointed out earlier, eased navigation through clear formulations and layout helps participants direct their full attention to the questions themselves. This principle can also be applied to the use of an online questionnaire with young-old participants who have low digital literacy. If they are not intimidated by taking a survey in an unfamiliar medium, navigation through the online questionnaire will also distract young-old participants from the actual questions. In conclusion, presenting the questionnaire on paper increased the number of responses and the quality of answers provided by young-old participants. 5.3 Administration: Sampling and Ethics While the sampling strategy (convenience sampling) for the pilot study was more direct and easier to accomplish, accessing the field (Legutke & Schramm 2016) and sampling of the main study were more complex. This section covers how vhs centres in Hesse (see chapter 4: research context) were contacted in order to find teachers and their learners interested in participating in the study. In a first phase, I contacted the central vhs of Hesse (Hessischer Volkshochschul‐ verband - hvv) to inform them about my study, the questionnaire, and the col‐ lection procedure, in order to elicit their support. Going through this institution also entailed a second phase: letters to the heads of language departments at all vhs centres in Hesse were sent out to raise the interest in my study. The letter also asked them to distribute a letter informing all English teachers at their institutions to contact me for further information and questionnaires. On the one hand, this was intended to conform to research ethics as teachers were approached indirectly while their contact details were protected. This re‐ assured teachers that their details were not given out to any researcher approa‐ ching their vhs centre from outside and they were not pressured into partici‐ pating. On the other hand, this meant that the formulation of the letter was even more crucial as it had to provide enough information about the study and en‐ courage teachers come forward and ask their older learners to become study participants (Dörnyei 2010: 65; Gay et al. 2012: 189 f.). 114 5. Numbers: The Survey I formulated two versions of the letter to the “gatekeepers” (Grum & Legutke 2016): one for the heads of the language departments at vhs centres, the other for the English teachers interested (see appendices B & C). The letter did not exceed one page and contained the official institutional letterhead with my con‐ tact details at the top, giving it a more professional look and thus potentially increasing the response rate. The same applied to the title behind my name at the bottom of the letter (foreign language teacher, diploma). The beginning of the letter included the title of the study giving the reader an immediate idea of the topic. In the first part of the letter, I created a connection between the institution vhs, their teachers and learners, and myself as a teacher and researcher. I per‐ sonalized the topic of my study by giving background information about my experiences of adult education and research in this field. I stated that from my experience the older language learner has received very little attention to date, and that it was the aim of my study to change that. This was intended to show both the institution and the teachers that my research interest was also relevant to their experience, since the majority of their learners are 50+ years old. In the second part, I gave the institution and teachers more information about who was doing the study and which other people and institutions supported it (su‐ pervisor and state vhs, hvv). This was followed by a detailed description of the questionnaire parts and length. Similar to the instructions of the questionnaire itself, it also assured them that it is not a test (i.e. there were no right or wrong answers) and highlighted the confidentiality of the data. I then gave the heads of departments and teachers further information or instructions on what to do next if they were interested in obtaining questionnaires for their learners (e.g. that they would be supplied with the necessary copies of the questionnaire and a return envelope, as well as a date by which the questionnaires needed to be returned). In the final part of the letter I once again thanked the heads of de‐ partments and teachers for their interest and help. I also emphasized in what way they and their learners would contribute to research in adult education and invited them to contact me by phone or e-mail if they had further questions. Shortly after sending out the letters via the hvv, I received e-mails from in‐ terested heads of departments and teachers requesting questionnaires, which were then sent out to them. The questionnaires were accompanied by a return envelope again increasing the potential return rate of questionnaires. In addition to this, teachers were also asked to complete a small form including the course level, the region or city, and the number of completed questionnaires. This in‐ formation was helpful later on when coding the questionnaires and correlating the results with these factors. As a result, 195 questionnaires were received from 115 5.3 Administration: Sampling and Ethics six different vhs centres in Hesse. The sample contains vhs centres from both urban (e.g. Frankfurt a. M.) and rural areas (e.g. Vogelsbergkreis). It also contains samples from all three administrative regions in Hesse: the northern (Witze‐ nhausen), central (Marburg, Marburg-Biedenkopf, Vogelsbergkreis), and sou‐ thern (Frankfurt a. M., Hanau) administrative regions. 5.4 Processing: Data Analysis This section focuses on the statistical processing or analysis of the completed questionnaires. Before the collected data could be processed using the software IBM SPSS © , it needed to be prepared for the input procedure by assigning codes to each questionnaire (i.e., participant). The codes contain information about the vhs centre, course level, and assigned a number to each participant. This made it possible to track irregularities in the data back to individual participant data. Since questionnaires were on paper, the data needed to be entered manually into SPSS. The following sections present the results of the data analysis using SPSS and are divided into three parts: 1. descriptive statistics (information about the demographics and motives of survey respondents), 2. inferential statistics of the Likert items in the first part of the questionn‐ aire, and 3. inferential statistics of the adjective scales in the second part of the qu‐ estionnaire. These sections on the survey results are followed by a discussion connected to the hypotheses formulated at the beginning of this chapter. 5.4.1 Descriptive Statistics: Demographics and Motives of Respondents This section looks at the demographics of the sample (195 respondents learning English as a foreign language at vhs centres in Hesse) and why respondents learn English. Since the above hypotheses are needed for this analysis, demo‐ graphic information regarding age distribution, retirement, time spent in Eng‐ lish classes at vhs centres, and gender will be looked at in the following. 116 5. Numbers: The Survey Age span <60 60-80 80> n/ s Number of all participants (= N; %) 33 (16.92) 153 (78.46) 4 (2.05) 5 (2.56) Table 5.2 Survey Participants: Age Spans My definition of the young-old with regard to the chronological age dimension states an age span from 60 to 80 (see chapter 2). As shown in table 5.2, the majority of the respondents in the survey are within that age span (153 parti‐ cipants; 78.46%). Only four are older than 80 years, with the oldest participant being 92. Of the 195, 33 (16.92%) are younger than 60, with the youngest parti‐ cipant being 45. As noted above regarding the intimacy of certain answers asked, all five participants who did not provide their age are female. But as emphasized in chapter 2, not only the chronological dimension of age will play a role in my analysis but also e.g. sociological dimension, which can be assessed through considering the circumstances in which young-old learners were born and grew up (cohort effects; generation), as well as their current role in society. In the case of my group of respondents, the factor of retirement strongly impacts the so‐ ciological dimension of their ageing. The descriptive statistics in this regard are as follows: Retirement working retired n/ s N (%) 28 (14.36) 163 (83.59) 4 (2.05) Table 5.3 Survey Participants: Working and Retired What will be important when trying to find an answer to the second hypothesis - “Being retired has a negative impact on L2 self-concept and its temporal as‐ pects.” - are the descriptive statistics about their working status: 163 (83.59%) participants are retired; 28 (14.36%) are not retired, and 4 (2.05%) did not tick an answer to this question. Another characteristic that sets young-old language learners apart from yo‐ unger age groups (e.g. students at school or young adults) is the often relatively long time they spend learning English at vhs centres. While children can expect to learn English as a foreign language for up to 13 years in the German school system, some young-old interviewees and survey respondents have spent se‐ veral decades learning English at the vhs centre. Therefore, this factor may play a significant role in shaping young-old learners’ self-concept. When it comes to 117 5.4 Processing: Data Analysis 5 Whenever I refer to the 60 to 80 group in the following overview and analysis, I am also including the four learners over 80 (two of them being 82, one being 84, and one being 92 years of age). how many years they have attended English classes at the vhs centre, I provided five time spans, this way making it easier for participants to simply tick where they fit in, instead of thinking about and calculating the semesters or years: Years spent at vhs centre <1 1-5 6-10 11-15 15> n/ s N (%) 48 (24.62) 97 (49.73) 23 (11.79) 14 (7.18) 9 (4.62) 4 (2.05) aged 60-80 30 (18.99) 83 (52.53) 21 (13.29) 12 (7.59) 8 (5.06) 1 (0.65) Table 5.4 Survey Participants: Time spent at vhs Centre Only four participants did not provide an answer to that question. 48 (24.62%) participants had spent relatively little time in their English class, namely less than a year (i.e. one or two semesters). The majority has already spent between one and five years learning English at the local vhs centre: 97 (49.73%). 23 (11.79%) participants spent six to ten, and 14 (7.18%) participants had spent 11 to 15 years at the vhs centre. Out of all 195 participants (indicated with N in the first row of the table above), 9 (4.62%) have been learning English for 15 or more years. Unfortunately, this does not give an idea of what the maximum amount of time spent at a vhs centre to learn English is. From the interviews with my own language learners, some have mentioned that they joined their English class already in the 1970s and 1980s. Some of them have only taken a break for an occasional one or two semesters due to such factors as illness or familial obli‐ gations. Thus, it is possible that some of those nine participants have spent 20 or 30 years attending English classes at their vhs centre. The numbers slightly change when the focus is reduced to the 158 participants between 60 and 80 (in the second row of the table above). 5 Compared to the distribution of all participants, here are approximately 5 % less participants bet‐ ween 60 and 80 years of age who have spent less than a year at their vhs centre (30; 18.99%). If we compare this to the 33 participants under 60 years of age, 54.55% of those 33 have been learning English at the vhs centre for less than a year. 36.36% of those 33 under 60, have spent one to five years at vhs centres. In all of the other, longer time spans, the 60to 80-year-olds display a slightly higher percentage than the sample as a whole (between 0.34 to 2.27% more than 118 5. Numbers: The Survey the whole group), with e.g. 5.06% having spent 15 or more years at their vhs centre (compared to 4.66% of the whole sample). Compared to these values, the under 60 group showed that only a few have been learning English at their vhs for more than 6 years (none in both 6to 10-years and more than 15-years ca‐ tegory; 6.06% in the 11 to 15-years category). Gender male female n/ s N (%) 41 (21.03) 152 (77.95) 2 (1) aged 60-80 31 (20) 124 (80) - Table 5.5 Survey Participants: Male and Female Finally, out of the 195 participants 41 (21.03%) are male and 152 (77.95%) female, and two participants did not indicate their gender (1 %). The values hardly change when reduced to the 60to 80-year-old group (male: 31 participants, 20 %; female: 124 participants, 80 %). To some extent, this also mirrors the “fe‐ minization of age” (see chapter 2). It also shows that language subjects at vhs centres are more popular among female participants than some other subject areas covered at vhs centres (see vhs statistics in chapter 5). A central finding within the descriptive statistics of this study, are the motives of young-old survey respondents for why they learn English at their vhs centre (question 1 in part II of the questionnaire). It provided participants with a range of answers to tick and additional space they could note down other reasons not anticipated by the researcher. In the case of this question, participants were also allowed to tick several answers. The top two reasons for learning English at a vhs centre among 60to 80-year-old participants (n=155) are memory (93.5%; 145) and travel (89 %; 138). Following this, the reasons “I enjoy learning together with other people” (61.3%; 95) and “I like the language.” (60 %; 93) were also favoured by a large number of participants. At the other end of the scale, very few participants indicated that they learned English at vhs for work purposes (4.5%; 7) or to help their children or grandchildren with their homework (12.3%; 19). 119 5.4 Processing: Data Analysis 145 138 95 93 68 62 60 46 31 28 19 7 0 20 40 60 80 100 120 140 160 Memory Travel Learning together with others I like the language Offer appealed to me Good experiences Enjoyment No other options in the area Friend/ partner Recommendation (Grand)child Work Motives for Learning English (60-80 years of age; n= 155) Figure 5.4 Motives for Learning English What is surprising about the top four reasons that more than half of the 155 young-old respondents between 60 and 80 years of age chose is the high im‐ portance they place on “memory” as it even surpasses the motive of learning English for travel purposes. This result appears to mirror the “active ageing” trend discussed in chapter 2, since many young-old intend to keep their minds active by learning a foreign language. It is a theme that has also appeared in the interviews in the following chapter and thus will need to be compared in the final merging stage in chapter 7. Nevertheless, the motive of learning English in order to be able to communicate when travelling is particularly strong among young-old language learners. If we connect this wish of being able to commu‐ nicate abroad to the discussion around possible selves in SLA research (see chapter 3.4.1), this motive constitutes a foundation for an ideal L2 self of young-old learners. The interviews in the next chapter will give more depth to this highly popular motif of young-old learners, by taking a look at interviewees’ experiences abroad (e.g. critical incidents in an L2 community) and their plans or visions in this regard. Another theme that is prominent in the interviews and mirrored here, is the social sphere of learning a language at an institution like the vhs centre. This is also apparent in the results of this question: 95 out of 155 young-old respondents learn English at the vhs centre because they enjoy lear‐ ning with others. Finally, more than half of the young-old survey respondents also expressed affective reasons by indicating that they learn the English lang‐ uage because they like it (93 out of 155). Some of the participants also took the 120 5. Numbers: The Survey opportunity to add further reasons for their language learning. These included, for example, refreshing or not forgetting their English skills, or communicating with friends and family overseas, or with neighbours from other countries. One participant (69 years of age) listed reasons such as a relaxed atmosphere of the vhs centre and its absence of pressure to attain certain grades, as well as enjoying the chance to discuss personal ideas and other topics with people of the same age. 5.4.2 Inferential Statistics: Multi-Item Likert Scales As already described above, the first part of the questionnaire consisted of 28 items formulated around different aspects of the L2 self and its temporal aspects. The items were grouped into seven different self-related multi-item parts, which were correlated with demographic information from the sample (e.g. gender, age etc.). The following list shows all seven multi-item groups (with the item numbers in the questionnaire in brackets; see also table 5.1 above) and a sample item (in italics) to illustrate the concept: 1. Past L2 Self (2, 6, 17, 26, 28) I always had a talent for learning foreign languages. (item 6) 2. Present L2 Self (1, 3, 4, 5, 8, 9, 12, 14, 15, 17, 18, 20, 21, 22, 23, 25 I am happy with my current level. (item 25) 3. Ideal/ Future L2 Self (10, 13, 16, 24) I can imagine a situation where I am speaking English with foreigners. (item 10) 4. General Self-Concept (2, 5, 7, 14, 20, 22, 25) I’m a typical foreign language learner. (item 22) 5. The perceived learning process (4, 8, 9, 12, 15, 21) I give up if I encounter problems when learning English. (item 15) 6. The perceived impact on feeling mentally more active, younger and fitter (18, 23) Learning English helps me to feel younger and fitter. (item 23) 7. The institution vhs centre and the course (1, 11, 27) My English class at the vhs centre meets my expectations of what constitutes an ideal foreign language instruction. (item 11) A first foundational test showed that the 28 items have a high reliability, Cron‐ bach’s a = 0.88. An overall self-concept score of all 28 items was calculated, i.e. the means of the responses to all items (MN = 3.99 SD = .565). Afterwards, the means for the multi-item groups were calculated for further testing of correla‐ 121 5.4 Processing: Data Analysis 6 I would like to thank Reinhard Bentrup ( Justus-Liebig University) and Dr. Ian Stephen (Macquarie University) for their advice and support with the software SPSS for the analysis of my data. tions and significances with the demographic information such as gender, age, years spent at the vhs centre etc. If the means of two groups were compared (e.g. male/ female or two of the age groups), an independent t-test was carried out for the individual multi-item groups above (Field 2009: 334; Gay et al. 2012: 351 f.). If the means of more than two groups were compared (e.g. the three age groups or the five different time spans participants had spent at a vhs centre learning English) an ANOVA test was carried out (Field 2009: 348; Gay et al. 2012: 357 f.). 6 The significance of differences was determined by the p-value and then evaluated by the effect size (r; w 2 ) of a significant difference (Field 2009: 56 ff.; Settinieri 2016: 336 f.). 1. Past L2 Self The test in this temporal self facet showed mostly no significant diffe‐ rences between the groups tested. Exceptions to this were the instances when factors such as gender and the two older age groups (61-70 and 71+) were compared. In terms of past self items, male participants show lower self-concept scores (MN = 3.63; SE = 0.09) than females (MN = 3.98; SE = 0.05). The difference is significant t(191) = -3.19, p < 0.05 (small to medium effect size: r = .20). When comparing the past self of the two age categories 61-70 and 70+, participants older than 70 display a higher self score (MN = 4.07; SE = .08) than the younger group (MN = 3.84; SE = .06). This dif‐ ference is significant with t(160) = -2.04, p < .05. However, with r = .02 the effect size (Field 2009: 56 ff.) is rather small. 2. Present L2 Self The test of comparing the means of Present L2 Self items revealed no significant differences in any of the groups compared. However, it may be worth mentioning that we find a tendency towards a significant dif‐ ference when comparing the two groups at the end of the scale: “time spent at vhs centre” (i.e. the group that has spent less than a year and the group that has spent more than 15 years learning English at a vhs centre, with the latter scoring higher: MN = 4.08 SE = 0.15). The test showed a p value of .052 (and a close to medium size effect of r = .26). 3. Ideal/ Future L2 Self ANOVA and t-tests of the ideal and future self items have only revealed significant differences in one area: “time spent at vhs centre”. The ANOVA tests revealed a significant difference in the ideal self between the five 122 5. Numbers: The Survey groups, who have spent different times at vhs centres learning English: F(20, 170) = 1.78, p < .05, w 2 = 0.08 (medium effect size). In other areas, for example, retired and non-retired groups, no significant difference were existent. When comparing participants in terms of their work-status, the non-retired group scored higher means (MN = 4.17, SE = 0.16) than the retired group (MN = 3.93, SE = 0.07). However, the difference is not sig‐ nificant: t(189) = -1.38, p > .05; the effect size is small (r = 0.1). 4. General (L2) Self-Concept Again, most tests in this area showed no significant differences. For example, in terms of gender, male participants showed only a slightly lower self-image mean (MN = 3.63, SE = 0.1) than female participants (MN = 3.76, SE = 0.06). This difference was not significant t(191) = -1.01, p > .05 and had a small effect size (r =.07). Significant differences in this compa‐ rison were found in the realm of the two groups who have spent very little time or a long time at the vhs centre learning English. Again the “more than 15 years”-group scored a higher mean (MN = 4.20; SE = 0.21) than the other group, which has spent less than a year at the vhs centre (MN = 3.64; SE = 0.1). The differences were significant with t(55) = -2.32, p < .05 with a medium effect size r = 0.30. 5. The perceived learning process Unlike the other tested groups, the ANOVA tests revealed a significant difference in the perceived learning process between the five groups, which have spent different times at vhs centres learning English: F(30, 160) = 1.517, p < .05, w 2 = 0.08 (medium effect size). In terms of retirement, no significant differences were discovered even though there are diffe‐ rences in means obtained with the non-retired participants scoring higher values (retired participants: MN = 3.81; SE = 0.06; non-retired participants: MN = 4.05; SE = 0.11). However, this difference is not significant t(189) = -1.67, p > .05; and has a small effect size (r = 0.12). 6. The perceived impact on feeling mentally more active, younger and fitter The tests in this comparatively small multi-item group, revealed no sig‐ nificant differences in any of the groups tested. Even when carrying out an ANOVA test with the different age groups, no significant differences were revealed: F(10, 183) = 1.227, p > .05. The same can be said about the retired/ non-retired groups. The retired group scored only a slightly higher mean value (MN = 4.1265, SE = 0.09) than the non-retired group (MN = 4.1250, SE = 0.18), the significance of this difference could not be confirmed: t(188) = 0.01, p > .05. 123 5.4 Processing: Data Analysis 7. The institution vhs centre and the course Similar to some of the results in the previous paragraphs, the only groups displaying a significant difference are in the “time spent at vhs centre” category. The t-tests revealed significant differences with the group of “less than a year” - having a lower mean value (MN = 4.30; SE = 0.1) than the group of “more than 15 years” (MN = 4.63; SE = 0.12). The dif‐ ferences were significant with t(55) = -1.39, p < .05; the effect size of this, however, is rather small (r = 0.18). 5.4.3 Inferential Statistics: Analysis of Adjective Scales As pointed out earlier (see section 5.2), question 3 was problematic, both for the use of opposing adjectives inhibiting the leeway of participants in their choice, as well as the formulation in the second part that contained two temporal con‐ cepts. While the negative or positive weight of the adjectives was kept in mind when looking at the results, only the first, clearly formulated part of question three will be presented in the following. Again, in order to find answers to the hypotheses postulated at the beginning of the chapter and listed below, demo‐ graphic factors such as age groups, retirement and non-retirement, gender, and time spent at vhs centre will be compared on eight different adjective scales: 1. Past and present L2-self are ranked higher than the future/ ideal L2 self among young-old learners. This discrepancy increases with age (i.e. when comparing different age groups among the young-old language learners). 2. Being retired has a negative impact on L2 self-concept and its temporal aspects. 3. The time spent in a vhs center English course has a positive effect on L2-self-concept. 4. Female language learners have a higher L2 self-concept than their male counterparts. In order to make differences clearer and reach approximately similar numbers of participants, I further combined the groups of age range as well as different times at vhs centres. In this way, the three age ranges were reduced to two groups: under and over 60 years of age. As for “times spent at vhs centre”, the five groups were reduced to three groups: less than one year, one to five years, and more than six years. 1. Active - Passive Taking a look at this scale and comparing the retired (R) and non-retired (NR) participants, there is a slightly stronger tendency towards “active” 124 5. Numbers: The Survey among the latter group. The female/ male comparison shows that female participants scored slightly higher at both extreme ends (active 1 - passive 5), whereas most male participants considered themselves rather active (2) in their current vhs centre English class. The comparison of partici‐ pants younger than 60 (u60) and older than 60 years of age (60+) shows a similar pattern, with younger participants scoring higher at both ends. But again, the difference is not big. A significantly clearer difference is shown in the comparison of the category “time spent at vhs centre”. In this case, the number of participants who have spent more than 6 years learning English at a vhs centre and indicated that they are “active” scored significantly higher values than the other two groups. 0% 10% 20% 30% 40% 50% 60% 1 2 3 4 5 active (1) passive (5) a (less than a year) a (1-5 years) a (more than 6 years) Figure 5.5 Adjective Scale: Active-Passive 2. Distracted - Focused While the retired and non-retired groups start off and end quite similarly on the scale, the retired (R) felt less “distracted” than the non-retired (NR) group. Interestingly, these proportions change when we look at the under and over 60 year olds. Participants younger than 60 replied that they feel more “focused” than the older group. The comparison of male and female participants shows a very similar curve, with female participants fee‐ ling “slightly more distracted”. The comparison in the time spent at vhs centre shows equally similar results. This time, however, the group that has spent less than a year at a vhs centre learning English scored slightly 125 5.4 Processing: Data Analysis higher scores, ahead of the groups “more than 6 years” and “1-5 years” on the “focused” side of the scale. 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 1 2 3 4 5 distracted (1) focused (5) a (NR) a (R) Figure 5.6 Adjective Scale: Distracted-Focused 3. Inhibited - Relaxed Even though the non-retired group has a higher score at the “relaxed” end of this scale, the retired group shows an overall stronger tendency on the “relaxed” side and a slightly lower score on the “inhibited” side of the scale. Things look a little different in the younger and older age group, with the 60+ group having highest scores on the “relaxed” side, the yo‐ unger group, however, displaying an overall higher “relaxed” score. When doing a gender comparison on this scale, both genders only show slight differences, with female participants displaying a stronger feeling of in‐ hibition. Again we can see a clear difference in scores of the group that has spent more than 6 years at a vhs centre compared to the others. This group has significantly higher scores on the “very relaxed” side of the scale (5). 126 5. Numbers: The Survey 0% 10% 20% 30% 40% 50% 60% 70% 1 2 3 4 5 inhibited (1) relaxed (5) a (less than a year) a (1-5 years) a (more than 6 years) Figure 5.7 Adjective Scale: Inhibited-Relaxed 4. Open - Careful towards new things Again, the retired and non-retired people have similar scores. The non-re‐ tired group, however, has higher scores on the “open towards new things” side of the scale. The reverse seems to be the case for the u60 and 60+ groups, with the latter scoring slightly higher on the “open towards new things” side but also on the other end of the scale, “careful towards new things”. In terms of gender, we see a gradually declining curve from “open” to “careful” for the female (F) participants, whereas the curve of the male (M) participants, peaks at 1 (open) and 3 of the scale. Once more, the group with the longest vhs centre experience has significantly higher scores on one end of the scale than the other two groups. It scores highly on the “open towards new things” side of the scale. This time, the difference is not as large as in the previous scales. 127 5.4 Processing: Data Analysis 0% 10% 20% 30% 40% 50% 60% 1 2 3 4 5 open (1) careful (5) towards new things a (M) a (F) Figure 5.8 Adjective Scale: Open-Careful Towards New Things 5. Strong - Weak On this scale, all four categories investigated show a similar pattern in participants’ answers. While the extreme ends - “strong” (1) and “weak” (5) - are comparatively low, the scores peak at the centre (3). The retired, participants under 60, male participants, and participants with the longest experience at their vhs centres show slightly higher values at the “strong” side of the scale and slightly lower scores on the “weak” side of the scale. 128 5. Numbers: The Survey 0% 10% 20% 30% 40% 50% 60% 1 2 3 4 5 strong (1) weak (5) a (NR) a (R) Figure 5.9 Adjective Scale: Strong-Weak (NR/ R) 0% 10% 20% 30% 40% 50% 60% 70% 1 2 3 4 5 strong (1) weak (5) a (u60) a (60+) Figure 5.10 Adjective Scale: Strong-Weak (u60/ 60+) 6. Cooperative - Independent The retired and non-retired score relatively similarly on this scale as well. The retired group, however, shows stronger peaks at 2 (“relatively coo‐ perative”) and 4 (“relatively independent”). The participants under 60 dis‐ play overall higher cooperative values than the 60+ group. While the fe‐ 129 5.4 Processing: Data Analysis male participants score higher values on both ends of this scale, the male participants show an overall higher cooperative score. Finally, the two groups who have spent either less than a year or less than six years at a vhs centre show a similar distribution of scores compared to the previous categories. The “6 or more years”-group, however, stands out again, sho‐ wing higher values at 1 (“cooperative”), 3, and 5 (“independent”) on the scale, but also significantly lower scores on 2 compared to the other two groups. It should be added that even though both adjective endings can be regarded as equally positive, all participants had a stronger tendency towards the cooperative side of the scale. 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 1 2 3 4 5 cooperative (1) independent (5) a (u60) a (60+) Figure 5.11 Adjective Scale: Cooperative-Independent (u60/ 60+) 130 5. Numbers: The Survey 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 1 2 3 4 5 cooperative (1) independent (5) a (M) a (F) Figure 5.12 Adjective Scale: Cooperative-Independent (M/ F) 7. Content - Ambitious The non-retired group shows higher scores - particularly at the “content” side - than the retired group. Both groups peak at 3 and 4 (undecided; “re‐ latively ambitious”), but with the retired participants scoring higher. The picture is again different for the u60 and 60+ groups. Here, the 60+ group scores higher on the “content” side of the scale (but not on the “ambitious” side). Both groups - similar to the retired/ non-retired category - peak at 3 and 4, but with participants under 60 scoring higher. The gender cate‐ gory is very similar to the retired/ non-retired group, with male partici‐ pants scoring higher at the 3/ 4 peak and women scoring higher on the extreme ends. Unlike before, it is the “less than a year” that stands out on this scale. The participants in this group reach significantly higher scores on the “ambitious” side of the scale (4 and 5). Both “1-5” and “more than 6 years” groups score higher values than the “less than a year” group on the “content” side. 131 5.4 Processing: Data Analysis 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 1 2 3 4 5 content (1) ambitious (5) a (less than a year) a (1-5 years) a (more than 6 years) Figure 5.13 Adjective Scale: Content-Ambitious 8. Introverted - Extroverted Similar to the strong/ weak scale, the scores of all participants in all cate‐ gories show similar curves with a peak in the centre (3). In terms of reti‐ rement, the retired group indicates higher inhibition than the non-retired group. In the comparison of the reverse is again the case, with the 60+ group showing higher extroversion scores. While male and female par‐ ticipants are very similar on the “introverted” side of the scale, female participant reach higher on extroversion. The “6 or more years” group shows significantly higher extroversion score than the other two groups. It also scores lowest on introversion. The “less than a year” group scores lowest on extroversion and highest on the introversion end of the scale. 132 5. Numbers: The Survey 0% 10% 20% 30% 40% 50% 60% 1 2 3 4 5 introverted (1) extroverted (5) a (less than a year) a (1-5 years) a (more than 6 years) Figure 5.14 Adjective Scale: Introverted-Extroverted 5.4.4 Discussion in Light of the Hypotheses In the following, the results of the previous two sections will help to find answers to the four hypotheses postulated at the beginning of the chapter. Both the re‐ sults of the multi-item Likert scales in Part I and - at least in terms of present L2 self or general self-concept - the results of question 3a in the second part of the questionnaire, serve as a basis of the discussion. Past and present L2-self scored higher than the future/ ideal L2-self among young-old learners. This discrepancy increases with age (i.e. when comparing dif‐ ferent age groups among the young-old language learners). The analyses of the Likert items for the past, present, and future/ ideal L2 self showed that even though the older age group 71+ scored higher means in every category, only the category past L2-self indicated a significant difference. While we need to be cautious regarding the results of question 3 - since the scores present a tendency rather than a calculated significance or effect size - we can tell from the results that participants over 60 tend to score higher when it comes to the feeling of being relaxed in class, extroverted, and content with their performance. They also tend to feel more distracted and they value cooperation more than inde‐ pendence. Overall then, there seems to be a difference in some present L2-self aspects if we split the age range in half, parting at 60. Based on the results of the Likert items tests, the hypothesis can only partly be supported: there is a diffe‐ rence in the past L2-self when comparing the three distinct age ranges. 133 5.4 Processing: Data Analysis Being retired has a negative impact on L2 self-concept and its temporal as‐ pects. As pointed out in chapter 2, the factor of retirement plays a crucial role in defining and understanding the young-old - particularly more than chronolo‐ gical age alone. In terms of the t-tests carried out for self-concept, as well as the past, present, and ideal/ future L2 self, the non-retired participants scored higher means than the retired group, except for slight differences in favour of the retired group in the realms of the past L2 self. However, all of these differences were not significant. Question 3a in the second part of the questionnaire, which gave an insight into their present L2 self, revealed that the retired group showed lower scores in terms of extroversion, openness towards new things, and activeness in class. They showed stronger tendencies in the areas of focus, cooperation, and strength. The other areas for both groups show similar tendencies towards feeling relaxed in class and ambition. All in all, and particularly in terms of the results of part I in the questionnaire, it can be concluded that this hypothesis cannot be fully supported by the data analysis results. The time spent in a vhs centre English course has a positive effect on L2-self-con‐ cept and its temporal facets. The ANOVA tests of all five groups have revealed significant differences in the areas of the ideal L2 self and perceived learning process. When it comes to comparing only the extreme ends or cases, i.e. participants who have spent less than a year and participants who have spent more than 15 years at a vhs centre learning English, we find significantly different scores in the areas of “general (L2) self-concept” and “the institution vhs centre and the course”. The t-tests of those two groups indicated an increased positive effect with time spent at the vhs centre. However, one can argue that this result is not necessarily surprising: Why would someone who is unhappy with the vhs centre or the course stay that long? We need to be careful when comparing these results with the dis‐ played results of question 3a in the earlier section. There, the five groups were “condensed” into three groups. Nevertheless, it is interesting to see that the group that has spent 6 or more years shows significantly higher scores at the end of the scales measuring extroversion, strength, openness towards new things, feeling relaxed in class, cooperation, and activeness. In the areas of focus and ambition, only participants who have spent less than a year at vhs centres show higher scores at these particular ends. Perhaps, again not a surprising result. From all the hypotheses, this one seems to receive the greatest support from the results in both the multi-item scale tests and the adjective scales in question 3a of part II. 134 5. Numbers: The Survey Female language learners have a higher L2-self-concept than their male coun‐ terparts. Similar to the answers to the first two hypotheses, there was only a significant difference in terms of past L2 self, with female participants scoring higher. For all the other tested concepts in the multi-item scales, no significant differences between male and female participants could be found. Nevertheless, if we take a look at question 3a again, female participants only show a higher score on the ends of being open towards new things, extroversion, and cooperation. Never‐ theless, on the whole, men and women reach similar scores on the adjective scales, thus, refuting this hypothesis. 5.5 Summary and Outlook The discussion of the results in light of the hypotheses in the previous section, draws a picture of the young-old respondents as being a relatively homogeneous group. There can be numerous reasons for finding little or no significant diffe‐ rences of young-old respondents, such as the sampling, their age or sociological age markers, and the questionnaire items. However, whether and why they are, in fact, a homogeneous group, as the above results suggest, or whether it is other factors that have yielded these results, is something to be further investigated by future research, namely research that includes other - younger or even older - age groups into the sample of foreign language learners at vhs centres (see chapter 8). Being confronted with the results of the questionnaire, its advantages and disadvantages again come to light. On the one hand, we can make clear statements within the scope of above the hypotheses and apply these to a greater population - the population being young-old English learners in the state of Hesse. On the other hand, this scope is very limited and leaves us with more questions, the main question being why? : Why did participants respond the way they did? Here, the qualitative data strand, i.e. the interviews, comes into play. As already discussed in the methodology chapter, combining the two data sets later on has several functions. What we need to keep in mind when looking at the data analysis of the qualitative data are mainly two things: Firstly, to what extent are the results of the qualitative data analysis in line with the ones pre‐ sented in this chapter? Secondly, to what extent can the qualitative data answer the “why? ” above? In other words, the next chapter will possibly add depth to the results in this chapter and to the answers of my research questions posed in chapter 4. By depth I mean, for instance, further exploring the categories of “time spent at vhs centre” and the “past L2 self ”, which have produced some significant 135 5.5 Summary and Outlook differences in the quantitative data analysis. The qualitative data analysis can possibly shed further light on why there are differences and what it is exactly that motivates participants to learn English for 15 or more years at a vhs centre. In what ways do those interview participants who are new to an English course differ from the ones who have already been a member of that group for several years or even decades? It can further enlighten the impact of the “past L2 self ” aspects on the English learning process. Of course, we also need to remember that both data strands are treated equally and that the interview questions in the qualitative part have only been influ‐ enced partly by the preliminary quantitative data analysis. It means that the qualitative data analysis can stand alone or go beyond the questionnaire scope - and vice versa. This is why most of the comparative and contrastive issues of both data strands will be explored in greater detail in the chapter on merging of this thesis (chapter 7). 136 5. Numbers: The Survey 1 This is in contrast to the concept of generalizability, which is a more suitable aim of larger-scale, quantitative studies such as the survey in the previous chapter. 6. Words: The Interviews Questionnaires might tell us that students hate drills, that they find them ‚boring’ or ‚pointless’, and their (recorded) behavior in class might confirm this, but only interviews can probe the beliefs and experiences that might explain their responses. In a profession like teaching, such understanding can be invaluable. - Richards 2009: 187 - Conducting interviews with language learners gives them the opportunity to express their views and provide in-depth insight into their perceptions of their language learner self and language learning experience (Mercer 2011a; Klippel 2011; Burwitz-Melzer & Steininger 2016). One aim of this interview study is to achieve transferability, i.e. yield “lessons learned” that can be transferred to other educational contexts (O’Leary 2014: 61; Nunan & Bailey 2009: 171 f.). 1 The detailed description of the research context (see chapter 4 and below), as well as the representation of all interview participants and the data analysis proce‐ dure (see appendix and below), constitute the foundation for making that transfer. Additionally, and particularly in the role of a qualitative strand within this mixed methods research project, the interview study aims to act as a coun‐ terweight for the quantitative data strand presented in the previous chapter. Richards’ quote above hints at this counterweight role as the interviews can further explain, expand, or contradict questionnaire responses (see chapter 7). With regard to the fundamentals of this research instrument, we need to keep in mind that an interview can be prone to social desirability effects (just like questionnaire responses). Further, the interviewer can more or less uncons‐ ciously influence interviewees’ responses simply by facial expressions, gestures, how questions are formulated etc. (Friebertshäuser 1997: 371; Murray 2009). Getting the questions and interaction right before, during, and after the inter‐ view are therefore the first crucial steps in an interview study, and will be dis‐ cussed in sections 6.1 and 6.2 below. In section 6.3, I present the data analysis of the vast amount of data collected. The qualitative content analysis has been chosen as a tool for reducing the amount of data and to make thematic patterns and clusters visible in the data (Kuckartz 2014; Mayring 2014; Schreier 2012). This section also presents the results of the complex data analysis procedure including interview summaries and discussion of the thematic categories using selected quotes from the interviews. 6.1 Controlled Flexibility? Approaching a Semi-Structured Interview I chose to conduct semi-structured interviews: With this type of interview, an interview guide is used that provides a rough structure for the interview while still allowing natural conversation. The main purpose of a well-constructed in‐ terview guide is to help the interviewer navigate smoothly through the inter‐ view, but it also helps the interviewee find and articulate answers to the ques‐ tions. It provides the scaffolding for the conversation while still leaving room for relevant questions that were not anticipated by the researcher. This is not to say that study participants take part in a mere question-and-answer game but rather in “jointly constructed encounters” (Richards 2009: 190; see also: Roulston 2014; Klippel 2011). Furthermore, the interview guide makes individual inter‐ views more comparable and lays the groundwork for category formation in the later stage of qualitative content analysis (see section 6.4; Schreier 2014; Schart 2003: 60, Bortz et al. 2006: 314). Connected to this, Friebertshäuser (1997: 375) sees a tendency of using semi-structured interviews in combination with other (quantitative) methods - which makes this type of interview even more valuable for my Mixed Method Research (MMR) framework. For the construction of the interview guide, I turned to different sources: a review of the literature, notes from classroom observation, the preliminary qu‐ estionnaire analysis (see chapter 5), and the pilot study of the interview (for refining and expanding the first draft of the interview guide). The interview questions aimed at operationalizing the theoretical considerations presented in chapters 2 and 3, or in other words, at eliciting self-perceptions of a young-old’s past, present, and future language learning process and the language learner self. Some of these questions elicited this in a more direct fashion (e.g. “Using three words - e.g. adjectives - how would you describe yourself as a person and as an English learner? ”), others did so in a more indirect way (e.g. “What kind of person should a new participant be in order to be a good fit for your course and to succeed at language learning? ”). The interview itself was accompanied by a language learning biography worksheet which the participants had been asked to complete before the interview. However, not all the participants had the time to do so (see 6.1.2). 138 6. Words: The Interviews 6.1.1 Overall Structure of the Interview Guide The interview guide (appendix D) is divided into four parts: preamble (or ope‐ ning), warm-up, main part, and closing (Reinders 2011b: 89 f.). In the first ver‐ sion, which was used in the pilot study, the preamble consisted of a paragraph introducing participants to the aims of the study. In the final version, which was used in the main study, the opening part was reduced to key words that were clear and sufficient for me as an interviewer to cover all key aspects of that introduction and speak freely, rather than read the text out loud. The four key points acted as reminders to provide the participants with information about the aims of the study and research project, to explain how the data (the recor‐ ding, transcript, language learner task) is handled, and other information re‐ garding confidentiality (e.g. the use of codes for each participant). Furthermore, the overall process of the interview was explained to the participants, i.e. that there were no correct or incorrect answers, and that if they could not answer a question they could skip it and bring it up again at the end of the interview (which was indicated by circling the question mark in the second column of the interview guide). This made the preamble quite similar in terms of contents and aims to the opening section of the questionnaire. Following the preamble is a warm-up section, which includes the question about why participants had taken up English classes at the vhs centre. This kind of question is easy to answer for participants and thus makes them more com‐ fortable in, what is to them an unusual interview situation, and, in this way, helps to initiate a more freely flowing conversation (Richards 2009: 188). The main part that follows the warm-up section is also the longest part of the in‐ terview guide containing all central questions, moving from questions addres‐ sing their past learning experiences, to their present perceptions of their lang‐ uage learning progress, and finishing off with questions eliciting a future outlook on their next semesters at the vhs centre and language learning plans beyond this context. Lastly, the interview guide ended in a closing part, in which I gave the participants the opportunity to add anything I may have not consi‐ dered when constructing the interview guide (Reinders 2011b: 92; O’Leary 2014: 228). Additionally, in the closing part of the interview, participants could ask the interviewer questions: this aimed at reversing roles and balancing the asking vs. being asked-situation and at aiding the leaving of the field. The question sections consist of three columns: the first column contains the interview questions. Each question section of the opening section, main part, and closing section contained a main question plus prompts and/ or probes (O’‐ Leary 2014: 228). The prompts provided the interviewer with an alternative for‐ mulation of the question in case the participant struggled to understand or find 139 6.1 Controlled Flexibility? Approaching a Semi-Structured Interview answers to the initial question. The probes aimed at delving deeper into the topic or extending it. The second column contains a question mark (in case a question was skipped and brought up again at the end of the interview because the par‐ ticipant could not find an answer), and a box to tick once a question had been answered (see appendix D). Conducting an interview is a multi-tasking ende‐ avour: asking questions, keeping an eye on the equipment and time while acti‐ vely listening to the responses of a study participant, and thinking about - based on what either the interview guide or the response have to offer - what to ask next. Thus, this second column made multi-tasking during the interview easier since ticking or circling took less time than making notes and it acted as a visual aid to follow up on the progress and to identify remaining unanswered ques‐ tions. Finally, the third column offered some space for additional note-taking for each question section. Similarly, there was space for note-taking after the interview (after a final thank you to the participant). 6.1.2 The Language Learning Biography I created a language learning biography worksheet as an interview aid for study participants. The worksheet consists of two parts (see appendix F for learning biographies of learners presented in 6.4.2): The first part lists a set of open qu‐ estions asking participants to remember when they started to learn English and how they progressed (or not) over the years. In the second part of the language learning biography, participants are asked to illustrate their progress with its high and low points on a graph (e.g. perceived competency in the language, confidence, moments of success or failure). They are also asked to choose three to four highs or lows in particular from their language learning history and comment on these at the bottom of the graph (year and notes). A sample graph in the instructions gave participants ideas on how to complete the graph. The purpose of this language learning worksheet is twofold: Firstly, it pre‐ pares participants for some of the key questions of the interview and they can do so in their own time. This gives them the opportunity to possibly look up old material and documents from English classes they have attended in the past, and to remember their language learning experiences more clearly. Secondly, having the completed worksheet in front of them during the interview gave participants something to both physically and mentally hold on to and refer back to. Since this worksheet was addressed in the first question of the main part immediately after the opening question, it eased interviewees into this unusual interview situation and helped them be more comfortable. 140 6. Words: The Interviews Not all of the participants, however, completed this worksheet. In these cases, we either talked about the questions of the worksheet in more general terms and in a more spontaneous fashion (some interview questions were similar to the ones on the worksheet). This, in turn, meant that, due to the more sponta‐ neous situation, participants’ answers may have become more superficial com‐ pared to the answers of those who were better prepared by completing the worksheet. 6.1.3 The Pilot Study While a pilot study cannot provide us with all the experience we need to be expert re‐ searchers, it can save a lot of time and energy by alerting us to potential problems that can be worked out before we begin the actual study. - Murray 2009: 49 f. - Similar to the questionnaire construction, the pilot study for the interviews also played a considerably important role in testing the questions and further refi‐ ning the interview guide (Murray 2009: 49 f.). Reinders (2011b: 94) refers to this process of interview construction as an inductive approach (in contrast to a deductive approach, i.e. formulating questions based on, for instance, the lite‐ rature review or the research questions). The selection of pilot study participants was driven by the aim of finding learners who are comparable in their demo‐ graphics (e.g. age range, retired, and language learning level or experience) to the learners interviewed in the main study (Gay et al. 2012: 387). Participants from both the pilot and main study attend English classes offered by the same vhs centre, but they are not in the same English classes. Thus, demographics of pilot study participants were similar to the ones of the main study participants: the age range of participants (60 to 80 years of age) and the fact that they are attending English classes at the same vhs centre, at the same level courses (B1 and B2 with a focus on conversation). The pilot study was conducted in June 2014 and involved interviewing four young-old English learners from an upper-intermediate English class (CEFR-level B2) with a focus on conversation. As I knew their English teacher and the class because I had substituted for their teacher a few times, it was not difficult to gain access to the field and find four volunteers for my pilot study interviews. I had the opportunity to present my study and its aims to the whole class at the beginning of one session in the semester. Participants had two weeks to think about whether they would like to participate. I was invited to their final session, which was an end-of-semester BBQ-event in a participant’s garden, where four volunteers approached me about the pilot study. I provided them 141 6.1 Controlled Flexibility? Approaching a Semi-Structured Interview 2 Author’s translation from German: „Welche Eigenschaften müsste deiner Meinung nach jemand, der Englisch lernen will, zum Unterricht mitbringen? “ with an informed consent form which contained more information on the study, my contact details, and, more importantly, their rights as participants (appendix E). I arranged interview sessions with the four volunteers in the days after. Friebertshäuser (1997: 390 f.) notes that the way an interview is conducted (e.g. the formality, the degree of structure) is also dependent on the target group being interviewed (see also O’Leary 2014: 218). Even though I would have been able to organize a meeting in a more formal context - i.e. a room at university or at the local vhs centre - almost all participants, in both the pilot and the main study, preferred to either meet at their homes or in a more informal context such as a café, restaurant, or pub. Only three found it more convenient to meet in the classroom after class. Reinders (2011b: 93) reports that conducting interviews in their ‘natural’ environment increases the motivation of interview participants. However, it also needs to be noted here, that informal contexts such as cafés, restaurants, or pubs were more prone to interruptions by waiters and made transcription of the audio material more difficult due to increased background noise. When comparing the guide for the pilot study with the final draft for the main interview study, the changes became apparent. In contrast to the former draft of the interview guide, I mainly added prompts or probes to the main questions that seemed rather difficult for several of the pilot study participants to answer. As already mentioned, the prompts and probes were formulated to make it easier for participants to grasp a language learning memory or to find answers to a question by making it more concrete. This sometimes meant to elicit a concept more indirectly. For example, I was interested in what characteristics they con‐ sidered important to successfully learn English or a foreign language in general, and which of these characteristics they thought they possessed. Thus, the qu‐ estion “Which characteristics or traits do you think someone who would like to learn English needs? ” 2 was followed by differently formulated versions of this questions (prompts, e.g.: “What makes a good or successful English learner? ”). This question was also followed by probes which aimed at eliciting participants’ thoughts in this area in a different, often more direct, way (e.g. “Which of these characteristics or traits do you think you possess? ”, “Which of these characte‐ ristics or traits would you like to possess or would you like to have more of ? ”) or in a more indirect way (e.g. “Do you have role models in class? ”). Another important aspect of conducting a pilot interview study that is often overlooked, is the opportunity to test the equipment. Interestingly, the equip‐ 142 6. Words: The Interviews ment I used was not initially apparent as such to the participants or people in the surroundings (e.g. waiters in a café). Instead of using a ‘classic’ recording device, I used both a mobile phone and a tablet to record the interviews. This way, I had two recordings in case one device failed to record or provided poor quality. Using two devices, I could also position one of them closer to the par‐ ticipant or to me. Participants were also already familiar with these two devices, as I had been using them in English classes every week for teaching purposes. Moreover, mobile phones on tabletops during conversation have become a fairly common sight in our digital age. I would argue that due to all of this, participants would more easily forget that they are being recorded and feel more relaxed during the interview than with the more traditional recording devices. Moreover, the pilot study helped me to practice and further refine my inter‐ view skills (Richards 2009: 189 f.; O’Leary 2014: 223). As a teacher with several years of experience in managing classroom interaction, I had acquired the skills of asking my learners questions, actively listening to their responses, and taking up things they said for further discussion in the classroom. Of course, this ex‐ perience prepared me for a one-on-one interview situation only to a certain degree. Conducting the four pilot interviews realigned, as well as extended, my initial classroom management skills for this new one-on-one interview situa‐ tion. 6.1.4 Re-Connecting to Theory: The Interview Questions Similar to the questionnaire, a plethora of potential interview questions needed to be reduced to the most essential ones that would help to find answers to the research questions. Reducing the number of interview questions also meant that the amount of time for the interview did not exceed an hour as promised to the participant in the informed consent. In the following, I will present the interview questions from the opening to the closing part of the interview guide’s final version (thus, containing both deductively and inductively collected questions). Furthermore, the interview questions were intended to target one or more of the L2-self theoretical concepts outlined in chapter 3 and in line with the multi-item scales of the survey: past, actual, ideal/ ought-to L2 self and the L2 learning environment (see table 6.1 below). Thus, the interview roughly progr‐ esses chronologically from looking at the L2 self in the past, then the present L2 self, and finally the future L2 self-guides of language learners participating in the interview study. 143 6.1 Controlled Flexibility? Approaching a Semi-Structured Interview 3 Interview questions in the original guide (see appendix) were not numbered. The num‐ bering also does not mirror or convey the hierarchical structure of the main initiating questions and their corresponding prompts or probes. The closing questions are not listed here. Interview questions 3 Targeted theoretical concepts Past L2 Self Actual L2 Self Ideal / Ought-to L2 Self L2 lear‐ ning en‐ viron‐ ment Why did you take up English at the vhs centre? X What was it like to be back at “school”? X How do you motivate yourself to keep going? / How have you motivated yourself over the years? X X What impacts your motivation in a positive/ negative way? X What do you expect from an English class at your vhs centre (in contrast to English classes you had previously / at school)? X Could you tell me more about your English language learning biography? X X Has anything with regard to how you approach learning English changed after your retirement? X X Could you tell me about a particularly pleasant / critical / positive / negative English class / activity / situation in class at the vhs centre? X Has the way you approach learning English changed over the years (e.g. compared to learning English at school)? If so, how? X X What kinds of characteristics or traits does someone need to bring to Eng‐ lish class? What makes a successful English language learner? X 144 6. Words: The Interviews Which of these ideal characteristics do you possess? X Do you have role-models in your class? What do you admire or like about them with regard to learning English? X X If you had magical powers, what cha‐ racteristics would you give yourself to improve your English language lear‐ ning? X How would you describe yourself as a person / a foreign language learner? X What role do you play in your class (if it was e.g. a football team)? X What role does the teacher play for learning English? (expectations, ex‐ periences, feedback/ corrections) X X How happy are you with your current English learning level? How would you rate it on a scale from 1 (very poor) to 10 (excellent) in general / in terms of skills, grammar, vocab? X How confident do you feel when using English a) in class b) abroad? X X How do you envision your time in the English class in the upcoming semes‐ ters? Do you have any goals or plans for the next semesters with regard to your English learning? X Would you like to change anything about the way you learn English in the future? What would you like to do more / less often or intensively inside or outside class? X Regarding learning English, what kind of advice would you give a po‐ tential new participant joining your class next semester? X What should a new participant be like in order to feel comfortable in your class or to be a “good fit” for your group? X X X Table 6.1 Interview Questions and their Theoretical Target(s). 145 6.1 Controlled Flexibility? Approaching a Semi-Structured Interview As already mentioned above, the opening section (table 6.1: questions 1 to 5) in the interview guide aimed at making the participant feel comfortable in the interview situation and at initiating a conversation between interviewer and interviewee. The question “Why did you take up English at the vhs centre? ” (table 6.1: question 1) was regarded as a suitable opening question that every participant would be able to answer without much hesitation or difficulty; this is why this question has no prompts but only probes, which delve deeper into the topic area. The first three probes to that question included asking them when they started, what it was like to be “back at school” (table 6.1: question 3; par‐ ticularly if the participant is or used to be a teacher), and asking about how other people in their environment (i.e. significant others) reacted to this decision to learn a language at a vhs centre. The three probes following these (table 6.1: questions 3 to 5) extended the main question into the present by asking them how they have stayed motivated to continue, what influences their motivation in a positive or negative way, and what expectations they have regarding their vhs centre English course (in comparison to English classes at school). The opening section is then followed by the main part (table 6.1.: questions 6 to 22). The first main question addresses the language learning biography, asking participants to illustrate their language learning process in the past up until now, and to pinpoint some highs and lows in their language learning ex‐ perience to the present. I decided to use this question at the beginning of the main section to further ease participants into the interview. Even though many participants did not complete the worksheet and thus had to talk in a more impromptu fashion about the questions, it still gave both interviewer and in‐ terviewee some sense of security. Participants had something to hold on to (physically) and refer to while talking to me. The prompts aim at eliciting more specific aspects of, and reasons for, what they wrote and drew, by asking e.g. “At this point you drew a high/ low curve. Why? ” and “In case of a low point: What motivated you to keep going despite this low? ”. The last prompt asked partici‐ pants if they would now - after explaining their illustration to me - change the graph in any way. After some insights from the pilot study, I decided to add a probe to this first question in the main part: “Working life/ retirement: In what ways has this changed your language learning (at the vhs centre)? ” The next main question asked participants if they remembered a particularly nice/ good or critical/ bad English lesson (table 6.1: question 8). To make it easier for them to remember such a situation, the prompts to this question were re‐ formulated using the words “positive/ negative memories” and “upsetting/ amusing situations”. Another prompt asked for tasks or texts that they at first found unusual or funny. In order to go further into depth with this question, the 146 6. Words: The Interviews probes asked participants about how they felt in the situation or afterwards, what they were thinking, as well as how they dealt with the situation at the time or afterwards. This is followed by a question that had been derived from the preliminary questionnaire results (see chapter 5): “When you think about your past English learning experiences and compare this with your current situation: How has the way you have been learning English changed? ” (table 6.1: question 9). In the preliminary results of the questionnaire, the majority of young-old learners ag‐ reed with the similar statement that they have changed their ways of learning English. It is not possible, however, to further ask the survey participants why and how exactly they did so. Thus, I hoped to get in-depth answers to this qu‐ estion through the interview. The prompts, in this case, ask participants about their routines when learning English, or, if they have particular methods or strategies for studying the language at home. The following questions in the main part aim more directly at participants’ beliefs regarding the ideal language learner’s characteristics or traits and the perceptions they had about themselves as language learners (table 6.1: questions 10 to 13). Firstly, participants are asked on a more general level what traits or characteristics they think someone needs to possess to learn English: “Which characteristics or traits do you think someone who would like to learn English needs? ” Accordingly, the prompts provided a different formulation asking for what makes a good or successful English learner or participant in class. As al‐ ready outlined in the previous section, the probes to this question redirected the main question of a hypothetical ideal language learner to the interviewees themselves by asking them, for example, “Which of these characteristics or traits do you possess in your opinion? ”. The next main question (table 6.1: question 14) asked participants to describe themselves as a person in general and as a language learner using three words (ideally adjectives). Through the pilot study, I found that it became easier for interviewees to answer this question if I asked them to either a) imagine I was writing a portrait about them for an encyclopaedia or newspaper, or, b) imagine how their partner or friends would describe them in three words. Closely con‐ nected to this main question, I added another question after the pilot study as‐ king participants to define their role in class (table 6.1: question 15). To make it again easier for participants to come up with an answer to this question, I com‐ pared the class to a football team in which every player has a certain position or role. I asked the interviewee what their role and their contribution to the “team” was. 147 6.1 Controlled Flexibility? Approaching a Semi-Structured Interview Hoping it would reveal some insight into their language learning environ‐ ment and because it lends itself to a good transition, the next main question moved from describing the ideal language learner to the ideal language teacher (table 6.1: question 16). It asked interview study participants about how impor‐ tant an English teacher (past and current) is to them, the role he or she should play in class, their perception of feedback or corrections by the teacher, and if and how satisfied the teacher is/ was with their language learning skills or progress. The interview guide then transitions to asking the participants how satisfied they are with their language skills (table 6.1: question 17). As a prompt, partici‐ pants could rate this - as on the graph in the language learning biography - on a scale from 1 to 10. One of the two probes asked for a further differentiation of how they rate themselves in view of individual skills or language learning features (i.e. reading, writing, listening, speaking, vocabulary, and grammar). The other probe (table 6.1: question 18) was added after the pilot study: “How confident to you feel a) in class and b) when travelling? ” This question focused less on the temporal self but rather on the spatial language learner self in context of the classroom and outside the classroom. So far, the interview guide tackled questions mainly directed at the partici‐ pants’ past and present language learner selves, as well as the L2 learning envi‐ ronment. The last questions in the main part (table 6.1: questions 19 to 22) are directed at their future language learner self: “How do you picture your time in your English class for the upcoming semesters at the vhs centre? ” The probes expanded this main question by asking participants about their goals in regard to language learning, if and in what way they would like to change their ways of learning English (strategies, methods etc.), if they would like to do anything dif‐ ferently (less or more often) inand outside the classroom, and if there is any‐ thing they could do without in their English class. For the final question, participants had to imagine a new participant joining their English course in the next semester - in a way, this still has a future com‐ ponent but it also brings the conversation back to a more general level. The in‐ terviewee was asked to think about any advice he or she would give a new Eng‐ lish class member, and also asked to comment on what kind of person a new course member would have to be, in order to be a good “fit” for the course or to feel comfortable in this particular class. Answers to this question can mirror in what ways interviewees perceive their learner group as a whole, the types of learners or people within the class, and to some extent their self-perceptions when it comes to approaching language learning. The probe to this question extended this to a more general level by asking for advice with regard to language learning tips (strategies, motivation, procedures, courses, teachers etc.). 148 6. Words: The Interviews 4 This question was not included in table 6.1 as it did not target theoretical concepts. As already mentioned above, the interview guide closes with asking the parti‐ cipant to think about what he or she would like to add or if there were any other things the interviewee would like to discuss that I did not anticipate when const‐ ructing the interview guide. 4 In the closing part, questions the participant could not answer earlier were brought up again. Finally, the interviewer-interviewee situation was reversed by giving the participant the opportunity to ask the inter‐ viewer any questions (about the project or personal). 6.2 The Participants: Sampling, Ethics, and Data Collection At the end of a really successful interview the interviewer will at least have covered all the intended topics and the respondent will feel that they have participated in a ‚conversation with a purpose’. - Richards 2009: 186 - The sampling strategy of my study falls into the category of convenience samp‐ ling, as the interview study participants were learners I had been teaching for se‐ veral years. This comes with certain advantages and disadvantages. Relying on convenience sampling can be problematic, as the sample can limit generalizability of the results (Cohen et al. 2013: 155f.; Mayring 2014: 12). An advantage of this way of finding participants, however, was the trust I had built with my learners over the period of time I had been teaching them (for some one semester, for others almost five years). In other words, gaining access to the field as well as building rapport between the interviewees and myself have been eased due to my roles as both teacher and researcher (Cohen et al. 2013: 89). Being aware of this dual role be‐ comes especially important when approaching the conduct of the interview and even more so when approaching data analysis. The process of acquiring participants for my interview study began before the start of my last semester at the vhs centre (February to May/ June 2014): I informed the head of the language department about my research project and received per‐ mission to interview learners at the vhs centre. I then informed my classes halfway into the semester about the interview study and that I was looking for volunteers to participate in the study. Towards the end of the semester, in May 2014, I brought along copies of the informed consent for my study (see appendix E) and the lang‐ uage learner biography tasks for those learners, who wanted to participate in my study. I once more assured them that their participation was voluntary, and that not participating or withdrawing from it at a later stage would not disadvantage 149 6.2 The Participants: Sampling, Ethics, and Data Collection them in any way. The informed consent form also outlined this and other of the participants’ rights as well as general information about the study, protection of their privacy and handling of the recorded data (secure storage and destruction of raw data after 5 years). Some learners approached me immediately after class to make an appointment for the interview, while others asked me to contact them via phone once they had read everything and checked their calendars at home. The interviews of the main study were conducted after the pilot phase (see sec‐ tion 6.1.3) in June and July 2014 and later transcribed by myself (see section 6.3.1 for the transcription procedure in my study and section 6.3.3 for interview sum‐ maries of all participants). In order to ensure the confidentiality of participants, they were first given participant codes and, in a later data analysis stage, pseudo‐ nyms; other identifiers potentially threatening that confidentiality were replaced by either pseudonyms (e.g. names of relatives or friends) or abbreviations (e.g. place names) in the process of data transcription and analysis. In terms of the data collection process, many interview study participants wanted to conduct their interview either in their home, in cafés, or at a pub, as I have outlined earlier in this chapter. Due to time constraints, three of the partici‐ pants wanted to conduct the interview in the classroom before or after the English class. In two cases, another participant was present while the other was inter‐ viewed. In the first case, this was due to the fact that a married couple invited me to their home to interview them one after another. In the second case, two parti‐ cipants, who were close friends and only joined the course two semesters prior, wanted to combine a brunch and interviews at the home of one of the partici‐ pant. Before conducting interviews in the latter two cases, I feared that intervie‐ wees would influence each other’s responses to the interview questions. As outlined in section 6.1.3, I used two recording devices: a smart phone and a tablet with a recording app. According to Kuckartz (2014: 134), the disadvantage of audio-recordings in an interview situation - as opposed to taking notes only during an interview - is that participants feel uncomfortable being recorded, which in turn impacts the quality of their answers. He points out that, nevertheless, these effects can be reduced when participants get used to the situation and forget about the recording device. As argued earlier in this chapter, using a smart phone and tablet are less likely to be regarded as foreign objects in the participants’ everyday life (compared to a traditional recording device) and thus act as a camouflage for the recording. I hoped by using that this technology, participants would more quickly forget that they were being recorded. In table 6.2 below, I give an overview of the interview study participants, sho‐ wing their pseudonym and their participant codes, which also revealed their gender (M or F) and the course they attended (G1, G2 = upper-intermediate with 150 6. Words: The Interviews focus on conversation, CEFR-level B2; G3 = intermediate, CEFR-level B1). I also indicated the participants’ age, if they are still working (NR) or retired (R), in what year they first joined the English course, and how many years they have thus spent learning English at the vhs centre (for short interview summaries, see appendix H). Name (Pseudonym) Code Age Retired (R) Not retired (NR) Joined vhs class in… Years spent in vhs class Barbara G1TF01 59 NR 1982 32 Astrid G1TF02 52 NR 2012 2 Anne G1TF03 67 R 2013 1 Dorothea G1TF04 61 R 2013 1 Christine G1TF05 63 R 1986 28 Peter G1TM01 65 R 1982 32 Christian G1TM02 64 R 1992 22 Ulf G1TM03 61 NR 1986 28 Wolfram G1TM04 70 R 1994 20 Diana G2TF01 73 R 2002 12 Alexandra G2TF02 79 R 2001 13 Maria G2TF03 65 R 1986 28 Brigitte G2TF04 65 R 1988 26 Jennifer G2TF05 68 R 2013 1 Lena G2TF06 69 R 2008 6 Tanja G3TF01 72 R 2003 11 Katharina G3TF04 67 R 2008 6 Ruth G3TF06 64 R 2012 2 Katja G3TF07 75 R 2013 1 Birgit G3TF09 62 R 2009 5 Adam G3TM01 81 R 2004 10 Table 6.2 Interview Participants 151 6.2 The Participants: Sampling, Ethics, and Data Collection 5 By using the term “adequate”, I am referring to the commonly used term in German research literature, namely Gegenstandsangemessenheit, describing the fit of the me‐ thodological framework to the research aims, material, and questions by making prin‐ cipled decisions on its constituents and steps involved. In his overview of the basic principles in qualitative content analysis, Mayring (2010: 50; 2014: 40) refers to this as “object reference” and “appropriateness of method”. 6.3 QUAL Data Analysis: Principles and Procedures Words should be weighed, not counted. - Jewish Proverb - With completion of the data collection stage, a significant amount of data in the form of audio files and completed forms (autobiographical data and language learner tasks) had been accumulated. A qualitative data analysis was needed to systematically reduce and structure, or ‘tidy’ (Schart 2003: 59), the amount of data to relevant themes in the data. The data analysis method would need to help shed light on the complex and multi-layered language learner self and its temporal facets in the interview transcripts. Another goal was to produce results that could be used for a merging procedure with the quantitative data strand, without much adaptation and thus possible blurring of the original meaning or depth of the qualitative analysis and interpretation. As in a good game of chess, thinking ahead a few steps when planning the research design and individual steps within the research process is worthwhile. It is even more vital in a mixed methods design, as it culminates in the final merging process of the two complex research strands. The method that best combines these aspects is qualitative content analysis (Kuckartz 2012; Mayring 2010 & 2014; Schreier 2014b). While content analysis started out as a quantitative method to systematically analyse textual material, its qualitative version has also gained prominence in qualitative-oriented re‐ search over the last decades (Burwitz-Melzer & Steininger 2016; Dörnyei 2007: 245). Various versions of it have been developed and further refined in order to help researchers choose adequate 5 variants within the qualitative content ana‐ lysis framework. In her article, Schreier (2014a) describes this as a “toolbox” from which a researcher needs to employ those options among the individual steps of qualitative content analysis which are ideal (or, as mentioned above, adequate) for their particular research project. In SLA research, qualitative content analysis has proven its merits since it allows for a rich description of more complex or multilayered issues and texts related to language learning and classroom interactions (Burwitz-Melzer & Steininger 2016; Klippel 2011). Going back to the quote at the beginning of this 152 6. Words: The Interviews chapter, Richards (2009: 187) points to this potential of interviews to probe le‐ arners’ beliefs and experiences in the foreign language classroom. To achieve this, qualitative data analysis builds on the generation and use of categories by abstracting from specific passages in the data. Mayring (2014: 40) and Schreier (2014b: 170) consider categories and the formation of a category frame at the heart of qualitative content analysis, thus being one of its central characteristics. The category system helps to do the aforementioned “tidying” of the data and to unfold relationships or patterns in the interview data by dissecting and com‐ paring its components: similarities or common themes, contradictions, varying perspectives (Burwitz-Melzer & Steininger 2016). Moreover, the detailed desc‐ ription of how I have arrived at my category system (see section 6.3.3) also “con‐ tributes to the intersubjectivity of the procedure, helping to make it possible for others to reconstruct or repeat the analysis” (Mayring 2014: 40). How researchers arrive at a final category framework, however, can vary within this method. Kuckartz (2012), for instance, writes that category formation occurs on a theoretical-empirical scale, while Schreier (2012 & 2014b) uses the corresponding terms conceptand data-driven approaches. Depending on the research questions, the data types, and the setup of the study, a researcher has to decide where to position his or her data analysis on this scale: a theory-/ concept-driven (deductive), data-driven (inductive) or mixed approach (deduc‐ tive-inductive) should be applied (it needs to be noted, however, that a purely inductive and purely deductive approach is neither possible nor advisable). In other words, the aforementioned adequacy of the method for one’s research project forms the foundation for the decision-making process regarding to ca‐ tegory formation within qualitative content analysis. I discuss this process of category-formation within my study in section 6.3.2 below. 6.3.1 Preparing the Data for Analysis: Transcription I transcribed the 21 interviews using the transcription software f5 and a foot pedal. Even though this was a rather tedious task, it had the advantage of pro‐ viding me with a first in-depth look at the responses of participants. This is why e.g. Murray (2009: 50 f.) also sees interview transcription as part of the data analysis. Transcription involves the transfer of spoken language into written form and thus involves a reduction of the complexities in oral interaction (e.g. intonation, accents, pauses, overlaps). Various transcription conventions offer different degrees of expressing these complexities in the transcribed data, and thus make it necessary to make an informed choice based on the goals of data analysis, readability, and resources (i.e. time and money) available (Kowal & 153 6.3 QUAL Data Analysis: Principles and Procedures O’Connell 2013; Kuckartz 2014: 135 ff.; Dörnyei 2007: 248; Burwitz-Melzer & Steininger 2016; Gläser-Zikuda 2011). This data preparation step also helped me to apply an initial reduction of the material by leaving out most of the small talk at the beginning and end of in‐ terviews if they did not contain relevant information. It needs to be noted here, that in some small talk passages interviewees do reveal additional things about themselves. This seemed to happen particularly at the end of an interview, when respondents were given the opportunity to ask me a question or comment on how the interview went. To increase readability for the data analysis stage, I decided to reduce the complexity of the spoken interaction using a basic form of transcription. I used the transcription rules of Dresing et al. (2015) as a guideline for my own trans‐ cription (see appendix G for transcription examples and rules). I indicated over‐ laps between two speakers, emphasized words, and pauses, false starters, laughs, interruptions, or incomprehensible passages. A more detailed transcription of other linguistic features (e.g. intonation) was not necessary in the case of my project, since the focus of analysis was more on the verbal component rather than most of the prosodic (e.g. intonation, pitch) and paralinguistic (e.g. ge‐ stures) features of the interviews (Kowal & O’Connell 2013: 66). I smoothed out linguistic features such as dialects and contractions in German to further inc‐ rease readability for data analysis (Kowal & O’Connell 2013: 69; Dresing et al. 2015). 6.3.2 Category Formation: Organizing and Interpreting Data The category formation process of qualitative analysis requires a researcher to dissect a whole (the interviews) into its parts, relating them to each other and re-connecting them with the theoretical considerations and research questions that initiated the whole endeavour. Dresing et al. (2015: 39), for instance, com‐ pare this procedure to sorting holiday pictures into different boxes or assigning labels to them by finding common themes or patterns in the pictures (places, food, people, landscapes etc.). Similar to this example from everyday life, a re‐ searcher needs to first get an overview of the data, systematically reduce data to recurring topics and patterns in the data. As I have mentioned above, desc‐ ribing how I have arrived at the final coding frame is at the heart of my quali‐ tative content analysis and will be discussed in this section. For my category formation procedure and for analysis purposes, I have used the software f4ana‐ lyse since it promised a high compatibility with the transcription data and sui‐ table features for creating, assigning, and re-ordering codes as well as memos, 154 6. Words: The Interviews and for retrieving text passages for individuals or a group of codes/ participants. The coding system, coded passages, and memos were also later available for export to .rtfand .xls-formats. Table 6.3 shows the steps of my category formation process, namely the pro‐ cedure that worked best within my project in the end. In the first rounds of the category formation process, however, deductive, inductive, and mixed approa‐ ches were applied. Thus, it needs to be pointed out that category formation is to a great extent an iterative or cyclical process. As the table indicates, steps 1 to 4 involved several rounds of forming and testing categories (indicated by the P-symbol). Instead of viewing these as “attempts” that failed and then starting category formation processes anew, I used the word “round” here to emphasize the cyclical nature of category formation and to show that each round has been impacted to some extent by the previous one. The three different rounds also showed that certain categories recurred and thus indicated the progress of sa‐ turation (Schreier 2014: 14; Morse & Maddox 2013: 526). In the following, I will briefly discuss how each of the three rounds during category formation pro‐ gressed, why they did or did not work, and how they influenced the following rounds. Steps of Category Formation in my Data Analysis: Deductive-Inductive Pro‐ cedure (Content-Structuring Analysis) Steps Procedures Deductive ↻ Developing main categories from research questions, theo‐ retical background, survey results, and interview guide. Trial Coding Testing and revising main categories through trial coding of 20 % of the interview material (= 4 interviews). Text retrieval Text retrieval of passages coded with main categories. Inductive Further differentiating individual main categories using the retrieved passages (i.e. creating sub-categories) Main Coding Coding all of the interview material using the refined, final coding frame (main and sub-categories). Inter-Coder Agreement Check Having another researcher code 10 % of the interview ma‐ terial (= 2 interviews) and compare overlaps and discrepan‐ cies. Summary and Analysis Summary and analysis of the results along the main and sub-categories. Table 6.3 Category Formation Process 155 6.3 QUAL Data Analysis: Principles and Procedures 6.3.2.1 ↻ Round 1: Concept-Driven Category Formation With the research questions, theory, and quantitative results in mind, I started from a concept-driven (deductive) approach to form categories for a trial coding of 10-20% of all interviews. For this first round and all of the following rounds, I chose three to four interviews that represented different genders, ages (learners in their 60s, 70s, and 80s) the courses (i.e. different levels), and extreme cases (e.g. extremely optimistic and pessimistic cases were selected based on the transcription process and an initial close rea‐ ding). I firstly reduced the concepts of past, actual, and future (ideal/ ought-to) selves as well as language learning environment to two main categories based on an actual-target differentiation: “Actual L2 self” and “Future L2 self”. The category “Actual L2 self” contained all past experiences and language lear‐ ning environments up until the present, i.e. everything that fed into the ac‐ tual state of a language learner self. The other category “Future L2 self” con‐ tained everything that language learners envisioned for their language learning progress in upcoming semesters and years: ideal, ought-to, and feared L2 selves. I used these two main categories for an initial trial coding of the selected interviews, hoping to eventually further differentiate the coded and retrieved passages into the more specific concept-driven catego‐ ries addressed in the research questions and QUAN results section above. After the initial coding with the two main categories and a first coding at‐ tempt with differentiated categories, however, I noticed that by imposing a concept-driven category system to my data, other relevant phenomena in the data were overlooked or not accounted for. Clearly, the category system was too broad and could not elicit crucial points in the data. 6.3.2.2 ↻ Round 2: Data-Driven Category Formation In my second round to form and to trial code categories, I decided on a data-driven (inductive) approach. I hoped to make up for the previously overlooked phenomena in the interviews and make the data “speak for it‐ self” rather than imposing a rigid category system. While this approach helped to remain open for unanticipated and relevant categories that were used in for the final coding, it also resulted in a rather complex or messy category system with categories that, on the one hand, were interesting, yet, on the other hand, led away from my initial research interest. Thus, this round was missing a focus derived from theory and research questions that would streamline the coding procedure. This round of category formation resulted in 198 categories and subcategories that arose out of the data and 156 6. Words: The Interviews suggested categories from the previous round. Clearly, this plethora of cate‐ gories could not be used to code all interviews and needed to be tied back more strongly to the research interests. 6.3.2.3 ↻ Round 3: Mixed Procedure In the third and final round, I opted for a mixed procedure that combined conceptand data-driven elements. In order to code the text or data, and yet, not to be led away from the research questions as in the second round, I decided to firstly code along the interview questions. After all, the interview questions were based on theoretical considerations (see section 6.1.4), allo‐ wing for a more directed first trial coding with the research interests in mind. I retrieved the coded passages and further differentiated the original categories using the retrieved coded passages from the trial coding. This se‐ cond, inductive step was more effective than in the first deductive round. After this final round using a mixed procedure, I completed a final version of my category system that could be used for coding all 21 interviews (see next section). 6.4 Results of the Interview Study 6.4.1 The Category System As mentioned above, researchers working with the qualitative content ana‐ lysis view the category formation as the core of this data analysis proce‐ dure. In the context of my study, the result of my category formation pro‐ cess, i.e. the category system presented in the table below, constitutes therefore one of the central findings (Schreier 2012: 219). 157 6.4 Results of the Interview Study Main Cate‐ gory / Theore‐ tical back‐ ground Sub-Catego‐ ries Definition & Coding Rules Internal frame of Re‐ ference: Past & Actual L2 Self Synchronic Comparison of English skills with other domains (other subjects or languages). Remarks on talent or ap‐ titude for languages or maths. Diachronic Comparison or development of language learner self throughout time (progress, decline, change due to re‐ tirement, critical incidents in class or abroad in the past that changed perception of language learner self, L2 confidence, language learning attitude and beliefs). External Frame of Re‐ ference: Past & Actual L2 Self No compa‐ rison Refuses to make comparisons to others in class or makes no comparison in interview. Equal / Iden‐ tical Perceives other learners in class as equal in their per‐ formance or identifies with them or their level. I am better/ ahead of others Learner sees self ahead of others in class or in their community in terms of language skills or overall per‐ formance. Others are better/ ahead of me Learners perceive other learners as better or role-mo‐ dels in certain aspects of learning or using English. Connection Expresses the feeling of a connection to the group and considers this close connection as one reason why he or she has stayed motivated to attend classes over the years. Remarks on how he or she has become a member of the group or how the group has become a team over the years. Obligation Feels obliged to stay in the group or return every se‐ mester to prevent it from falling apart (e.g. obligation towards other participants, might lose interest or give up without them; obligation towards group or teacher, without them there would be too few learners to carry out planned learning activities or the group size would become too small to continue the course overall). Positioning & Role Talk about the role they play in class or how they po‐ sition themselves within the group (e.g. using a soccer team as a metaphor). 158 6. Words: The Interviews Main Cate‐ gory / Theore‐ tical back‐ ground Sub-Catego‐ ries Definition & Coding Rules Actual L2 Self Indepen‐ dence & Self-Deter‐ mination Awareness of a different power-relationship between adult learner and teacher (e.g. the role of a customer). With this comes a more self-determined way of lear‐ ning, attitude, and expectations towards the learning environment. Resilience Learner talks about how he or she is able or has learned to rebuild or to motivate self again after a setback or negative language learning experience. L2 Self-Con‐ fidence in-/ outside class‐ room Learners compare how confident they feel when using English in the classroom to when using it abroad with native or non-native speakers. Intended Ef‐ fort / Investment Degree of ambition or willingness to invest in language learning. Learners remark on how “lazy” or “ambi‐ tious” they perceive themselves as language learners and if/ to what extent this has changed over the years (e.g. before and after retirement). Vision: Ideal L2 Self strong Has an idea or clear vision of what he or she would still like or need to improve. Learner sees an expansion/ improvement of skills or reaching goals as still attai‐ nable/ possible. medium Learner has some ideas for improvement or goals, but does not expect to get very close to reaching them. He or she is happy or content with maintaining current level instead of further expanding English skills. weak Learner notices a decrease of language skills despite attending the class (e.g. forgets more and more, has become more inhibited, has lost passion). Has no ideas for improvement or goals (no vision). Perspective “we” Learner changes perspective using “we” (wir) to talk from the perspective of the whole English group. The learner does this to distance the “we as a group” from others (e.g. the teacher or learner) or to amplify an ar‐ gument. “you”/ ”one” Learner uses the word “you” (du) or “one” (man) to e.g. generalize an argument, speak hypothetically, or dis‐ tance self from a statement. Table 6.4 Category System 159 6.4 Results of the Interview Study 6 I would like thank Eva Schmidt for acting as a superb inter-coder by opening up new perspectives on the interview data. In order to test the reliability of the category system, an inter-coder agreement check was conducted (Mayring 2014; Kuckartz 2014: 167 f.; Burwitz-Melzer & Steininger 2016: 258; Gläser-Zikuda 2011). The online software QCAmap © was used for the inter-coder check since it enabled me to share the interview material with another researcher (inter-coder) in a confidential environment that has been specifically designed for qualitative content analysis and inter-coder checks. 6 Mayring (2014: 114) differentiates between three different levels of inter-coder agreement, depending on how much information for the coding process is revealed to the inter-coder. I provided the inter-coder with my cate‐ gory system and coding rules, definitions and coding rules and discussed my research objectives with her in detail so that she was familiar with my definitions and research questions. I did not provide the inter-coder with interview material that has already been coded by me in order to find out to what extent my cate‐ gory system and my coding procedure have taken effect. The inter-coder, who is familiar with this type of qualitative content analysis from her own research and the software, coded two interviews (i.e. 10 % of all interview material). The two interviews for the inter-coder check were chosen based on the characteristics of the interviewees (close to my definition of young-old in terms of chronological and sociological age; a male and a female participant from two different groups), and variation/ coverage of categories I used in my coding process (some degree of variation in the evaluative main category “vision”; covering all categories in order to test their reliability). After the inter-coder finished her coding of the two interviews, the inter-coder agreement check is closed manually and a comparison of coded pas‐ sages of both coders is made visible. Based on how well the chosen categories and the extent to which coded passages overlap, QCAmap allows the researcher to self-evaluate the inter-coder agreement on a scale (“excellent”, “good”, “mo‐ derate”, “bad”) and to add comments in regard to the decision making. After comparing the coded passages, the inter-coder agreement in my study was eva‐ luated to obtain a reliable (“good”) agreement of codings. The vast majority of codings overlapped between coders. In cases where there was disagreement, it could be ascribed to, firstly, a different approach to coding passages between the inter-coder and myself, i.e. how much of a relevant passage was included (for example, including or excluding interviewer questions). Secondly, another prob‐ 160 6. Words: The Interviews lematic area for the inter-coder agreement was the category “diachronic” as it was relatively broad compared to other categories, and thus led to some diver‐ gence between the two coders. The good result of the inter-coder agreement, shows how well the developed categories are able to grasp the recurring and relevant themes in the interview material. What follows in the next sections, are a brief presentation of interview summaries, and finally, a more detailed analysis of purposely selected participants. These results help to exemplify the aforementioned central finding of my interview study: the category system. 6.4.2 Presentation Sampling: L2-Vision in Focus, L2-Vision as Focus My use of the concept of imagination refers to a process of expanding our self by trans‐ cending our time and space and creating new images of the world and ourselves. Imagi‐ nation in this sense is looking at an apple seed and seeing a tree. It is playing scales on a piano, and envisioning a concert hall. - Wenger 1998: 176 - As suggested by Kuckartz (2012: 94), the results of the qualitative data analysis should not be presented in alphabetical order, or in the order of the original coding system depicted above, but rather in a logical and reader-friendly order. To achieve this, I will use the main category “L2-vision” to select purposively young-old language learners from my interview data pool, and create profiles based on how they position themselves within the whole category system. I chose the category L2-vision because of its evaluative nature (as will be outlined in detail below) and because it is a category which explores future self-guides (see chapter 3). Thus, it is hoped that grouping young-old learners based on their L2-vision can result in implications, such as powerful or effective motivational strategies for foreign language instructors and educational institutions. This way of presenting and discussing the outcomes of the qualitative data analysis does not only give insight into how young-old learners conceptualize them‐ selves as foreign language learners (main research question). It also has the potential to reveal what young-old learners with a strong L2-vision are doing or seeing differently compared to those with a weaker or even non-existent vision for their foreign language learning. Going back to the theoretical foundations outlined in chapter 3.4, the category definitions of L2-vision and its sub-categories capture the theoretical concepts of ideal L2 self (also referred to as a future self-guide together with ought-to L2 self). Wenger’s quote above reflects what is meant by the main category of L2 vision: interview study participants were asked to imagine themselves and their language learning journey in future semesters. Their responses indicated - not 161 6.4 Results of the Interview Study to what extent they would envision themselves playing in a concert hall - but rather how well they imagine their future L2 self inside and possibly outside the classroom, progressing and performing in the target language. Relevant pas‐ sages were coded depending on the strength or existence of their L2 self-image for their future (possible) self - e.g. how vivid or concrete are descriptions of their plans and goals for the next semester(s)? The quality of their L2-vision was also taken into consideration when coding the interview scripts. For instance, are learners optimistic about their language learning progress (strong L2-vi‐ sion)? Are their main intentions rather “humble” in that they aim to simply maintain their current language learning level (medium L2-vision)? Or is it that they are rather pessimistic based on previous or current language learning ex‐ perience and ageing so that their vision is either one of decline, leaving the class, or blurred/ non-existent (weak L2-vision)? To illustrate the concepts of main and sub-categories, table 6.6 below contains examples of coded passages for strong, medium, and weak L2 vision in the interview data. Since “L2-vision” is the only evaluative category (Kuckartz 2014: 44) within my category system, it lends itself particularly well to group young-old learners according to the strength of their L2-vision (sub-categories are strong, medium, and weak L2-vision), i.e. according to their outlook on their future or ideal L2-language learning self. Sorting learners based on their vision resulted in table 6.5 below, which gives an overview of all young-old learners in this interview study. The table indicates the number of coded passages of each sub-category (strong, medium, weak) which can be ascribed to each participant. Furthermore, it highlights non-retired participants and those who are older than 70 years of age. While the development of this particular main category has its roots in the theory discussed in chapter 3 (deductive), it aids the presentation sampling (cf. Legutke 2016) for the following sections in a data-based way (inductive). In other words, table 6.5 shows a data-generated continuum (weak to strong L2-vision) from which young-old learners can be selected and portrayed in the following. The portraits or profiles of the selected young-old language learners below help to illustrate the developed category system (see section 6.3.3), with categories showing varying manifestations within each participant’s interview. I chose six learners from the spectrum of weak to strong L2-vision. Further‐ more, learners with different characteristics were chosen that are related to the results of the inferential statistics in the previous chapter: the importance of time spent at the vhs, gender, different age spans within the 60 to 80 years of age span. This aimed at reaching a variety of perspectives on individual cate‐ gories, but it also aimed at laying a foundation for the merging of quantitative and qualitative data strands in the following chapter. Thus, three male (Adam, 162 6. Words: The Interviews Ulf, Wolfram) and three female learners (Anne, Dorothea, Tanja) are represented in the following. All of them except for Ulf are retired (I chose to represent Ulf as he was one of only two weak L2 vision learners). Adam, Wolfram, and Tanja are older than 70 years of age, the others between 60 and 69 years of age. Anne and Dorothea have been in their course for less than two years. Tanja and Adam have been in their vhs classes between 2 to 15 years. Wolfram and Ulf have been part of their English group for more than 15 years. 163 6.4 Results of the Interview Study Overview of L2-Vision Groups: Strong (S), Medium (M), Weak (W) (other characteristics indicated in table: 70+; not retired) S S+/ M M/ S M+/ S M M+/ W W/ M W+/ M W Anne (4S) Diana (3S) Tanja (3S) Ruth (3S) Lena (1S) Barbara (3S, 2M) Christine (3S, 1M, 1W) Peter (1S, 1M) Birgit (1S, 1M) Wolfram (6M, 2S) Astrid (4M, 2S) Jennifer (3M, 1S) Katja (2M, 2S, 1W) Christian (2M) Brigitte (2M) Dorothea (1M) Alexandra (5M, 1W) Katharina (3M, 1W) Maria (1M, 1W) Adam (7W, 5M) Ulf (4W, 1M) N/ A Table 6.5 Overview of L2-Vision Groups 164 6. Words: The Interviews 7 I: = Interviewer Examples for each L2-Vision Sub-Category Strong L2-Vision (S) Definition: Has an idea or clear vision of what he or she would still like or need to improve. Le‐ arner sees an expansion/ impro‐ vement of skills or reaching goals as still attainable/ possible. J: And that is basically this long-term goal: The hap‐ piness, I have when I’m in the country or when I have to use this language. That I will be able to use it fairly well. That is like a light back there somewhere, which keeps me motivated to study these languages again and again. Jennifer (G2TF05, section 204) Medium L2-Vision (M) Definition: Learner has some ideas for improvement or goals, but does not expect to get very close to reaching them. He or she is happy or content with maintaining current level in‐ stead of further expanding En‐ glish skills. I: 7 How happy are you with your current level in En‐ glish? C: Well, since I don’t have any expectations towards myself in that regard, I’m very happy with it. Christian (G1TM02, sections 229-230) Weak L2-Vision (W) Definition: Learner notices a decrease of language skills des‐ pite attending the class (e.g. for‐ gets more and more, has be‐ come more inhibited, has lost passion). Has no ideas for im‐ provement or goals (no vision). I: Maybe we should consider the future now: How do you picture your next semesters in the English class at the vhs centre? A: [knocks on table] That when you return in two years, I won’t be there anymore. I: Is that a plan of yours? A: It depends what our new teacher will be like and how that/ and because my pension is very low. Adam (G3TM01, sections 231-234) Table 6.6 Examples for L2-Vision Sub-Categories 165 6.4 Results of the Interview Study 8 Quotes from the interviews were translated by the author. I = Interviewer. For more transcription rules, see appendix. L2-Vision Groups and Time spent at vhs-centre (other characteristics indicated in table: Name S = intense S-Vision; Name W = intense W-Vision; Name = considered discontinuing course at time of interview) Less than 2 years 2 to 15 years More than 15 years S-Vision Anne S Ruth S Diana S Tanja S Lena S Barbara S Christine S M-Vision Dorothea Astrid Jennifer Katja Alexandra Katharina Brigitte Christian Wolfram W-Vision - Adam W Ulf W Table 6.7 L2-Vision Groups (Time at vhs Centre) 6.4.2.1 Weak L2-Vision Learners 6.4.2.1.1 “I don’t go there to learn something”: Adam Adam is the oldest interview participant and the only one over 80 years of age. His interview contains the highest number of weak L2-vision codings of all young-old learners. The retrieved passages with this coding show signs of in‐ difference with regard to the actual language learning. One of the factors that contribute to this is his financial situation: He has been struggling to pay for commuting to the vhs centre (approx. 20 minutes by car; Schmidt-Hertha 2014: 41) and the course fee with his small pension: Then I tell myself, well, when I turn 82 now, why should I still be doing this to myself financially? The driving costs money. The course costs money and all. Adam: 244  8 Another factor that appears within the category weak L2-vision (connected to this is also the category “diachronic”) is the biological dimension of ageing (see chapter 2) as he has noticed a decline in his overall cognitive abilities (memory). He also indicated this decline - regarding his language learning progress - in his learning biography task that we discussed in the interview: I: Okay, you wrote here, that you always forget a little bit, or tend to forget more. 166 6. Words: The Interviews A: Yes. More every day. It is getting worse. Adam: 83f It appears that this decline in memory over the last years, the financial burden, and the weekly commute have left their mark with regard to his outlook on future semesters. Like a few other language learners in their interviews he also appears to be sceptical or discouraged by a frequent change of teachers over the years and for the new semester. Being faced with a new teacher and a new teaching style to get accustomed to seems to give him an additional reason to think about disconti‐ nuing his English classes: I: Maybe we should consider the future now: How do you picture your next semesters in the English class at the vhs centre? A: [knocks on table] That, when you return in two years, I won’t be there anymore. I: Is that a plan of yours? A: It depends what our new teacher will be like and how that/ and because my pension is very low. Adam: 231-234 If you continued teaching us, then I would surely do one or two more courses. So‐ meday, I will probably become tired and the driving. And, what for? Adam: 254 His increased indifference or loss of passion about learning English also comes to bear when he is asked what kinds of topics he would like to discuss more often in future semesters in his vhs centre class. He mentions soccer (a current topic on the news and in his memory training class), but then adds that he would let it wash over him or endure whatever is discussed in class: A: Talking about some current affairs. About some (.)/ I: Okay. For instance? Soccer or politics? A: At the moment soccer, too. We talked about this in my memory training class. Other than that, no. Whatever is done in class, I let that wash over me. Adam: 273-276 In conjunction with this indifference or lack of ambition in regard to learning En‐ glish, he also mentions that he does not need the foreign language skills anymore as he stopped travelling. This connection to an L2-community where a learner can use his language skills is also an important aspect in the interviews of other young-old participants (both in a positive and negative sense). Adam shows a si‐ 167 6.4 Results of the Interview Study milar reaction of indifference to my question about any positive experiences or highlights in class over the past years (he joined the course around 2004): A: Actually, no. I: Nothing? A: Well, it’s stable. It’s just the same. Adam: 42-44 Adam also serves as an example for the dynamics of ideal and feared possible selves and how they can function as guides (see chapter 3.3.1), i.e. behaving in a way that drives a learner towards or away from a possible self. The biological factor mentioned above - Adam becoming more forgetful every day - is part of this dynamic that makes him continue courses such as English (among other lifelong learning opportunities such as IT courses and a special memory training class in a local hospital). He mentions his memory and not forgetting everything (about the English he once learnerd at school) at several instances during the interview. The feared self of losing his independence and the ability to leave the house due to dementia or Alzheimers manifests itself in people around him that are his age. The only other male English class participant, to whom he had a good connection, developed Alzheimers and had to leave the course two se‐ mesters earlier due to his illness (see below). Adam also sees himself ahead of others who are his age outside the classroom mirroring the coping strategies in old age observed in gerontological research (chapter 2.1.4; category: “external frame of reference - I’m ahead of others”; q.v. Anne below). But there are also positive aspects about the vhs centre English class that keep Adam coming back and attending the classes. The connection to the group is one example (category “external frame - connection”). Adam’s case exemplifies how having a change in everyday life and getting together with like-minded people young-old learners can socialize with are common themes within the category of “external frame of reference: connection”. The motivation to get out of the house and be among others with the same interests is particularly strong in the interviews of Adam. He told me that since he has difficulties with his wife at home, going to English class one morning a week is something he is parti‐ cularly looking forward to since in this way he can be away from his troubles (and once more mentioning his memory as a rationale): But ok, I wanted to be away from home once a week in the morning. Then it was nice company. I had a really good connection to Heinrich, who now / now that we aren’t in touch anymore. In the end he/ and well. It was also just to train my memory, because you forget more and more. 168 6. Words: The Interviews Adam: 40 He uses the same formulation of “nice company” in another instance later in the interview where he highlights more vehemently his main or actual intention for coming to English class: I don’t go there to learn something. I go there to be away [from home] and nice com‐ pany there and/ well. Adam: 152 When asked if he remembers a particularly nice English lesson he had in the last ten years he instantly replied that he liked the lunches we celebrated at the end of every semester. He explains that these lunches were always a “different kind of get-together” highlighting the socializing aspect of this English course: A: The lunch at the end of the semester. [laughs] I: [laughs] That’s something. Why this of all things? A: Because that was a different kind of get-together and well/ Adam: 100-102 Summarizing the case of Adam, a young-old learner with a weak L2-vision, the weighing of different purposes or rationales for learning English or attending foreign language classes is crucial. A feared possible self (not an L2 possible self) - that of losing independence in old age due to illnesses that impact cognitive abilities is one key motivator for Adam to continue lifelong learning. Another key factor that keeps Adam in this class is socializing, i.e. being part of a group that has grown together over several semesters or years (q.v. Tanja below), creating a routine away from home that is worth the money and the long com‐ mute. 6.4.2.1.2 Things are different as an Adult Learner: Ulf Ulf ’s interview also contains a high number of weak L2-vision codings. What Ulf and Adam appear to share, is the current and future lack of purpose in using English abroad or a lack of connection to the L2-community at the time of the interview (even though Ulf once took up English for a visit to the United States and he is passionate about country music). And just like Adam, he is thinking about discontinuing the course after the current semester. Adam and Ulf ’s lack of purpose or connection to an L2 community appears in the category “intended effort / investment”. In the analysis of this category two words have been used by almost all interview participants to describe the ideal language learner: “ambition” and “lazy”. For the learners, ambition seems 169 6.4 Results of the Interview Study to go hand in hand with a purpose or aim (e.g. a holiday, getting a certificate, or work), as well as an interest or passion for the language and culture. Interes‐ tingly, a few learners like Ulf create an ideal language learner by contrasting it with themselves as language learners, i.e. in the sense of “don’t do it like me”. U: Well. [sighs] Don’t do it like me. [laughs] I: Meaning? How should he do it? U: Well, develop more ambition. If he wants to progress, when he has the ambition [transl. lit: goal], then he simply should do something about it. Then you can certainly learn something from your course. These goals, they are not necessarily existent for me anymore. Ulf: 304-306 The goal or purpose theme reappears within Ulf ’s interview passages which were coded “intended effort/ investment” (a sub-category of “Actual L2 self ”). In line with Mercer’s (2011a) study on language learner self-concept, the ca‐ tegory “synchronic” focused on domain-specific aspects of young-old learners’ perceptions of self (see also: Shavelson et al. 1976: 413; Marsh et al. 1986; see chapter 3). Instances where learners made references to, and comparisons bet‐ ween, domains (i.e. other foreign languages and subjects such as maths or sci‐ ence) were coded this way. What became apparent in the summary and analysis of the coded passages was how participants viewed their general aptitude or talent for languages and the role of English besides other foreign languages such as French, Italian, Spanish etc. Interestingly, reflecting Marsh et al.’s findings (1986; see chapter 3: distinctiveness of maths and verbal academic self-concepts), several young-old learners interviewed claim that they have no talent or apti‐ tude for languages (Sprachenbegabter) in contrast to maths, physics or other science subjects they took at school. Some claimed that they were not particu‐ larly interested in learning English at school either due to it being a situation where they had no choice, or due to the teaching methods and teacher. Asking why they then decided to go back to learning a language even though they do not consider themselves as being good at it, some replied that they had a specific purpose or goal in refreshing/ learning the language, Ulf being a case in point: U: Chemistry and physics. These were my subjects at school. And mathematics. These are like/ I am more that type. I am not good with languages. I: And yet, you keep learning languages. U: I did it. Yes. But only, because I had a certain purpose. Or a certain goal. Ulf: 93-100 170 6. Words: The Interviews It is worth noting that, in line with the above analysis, Ulf uses the past tense when talking about his purpose or goal, further highlighting the weak tendency of his L2-vision. The lack of talent and preference for the subject English at school comes up earlier when we talk about his language learning biography: U: That was nothing. I wrote down “compulsory subject”. Yes, then it went/ I learned something there / then maybe I stagnated and than I restarted [at the vhs centre]. I also learned something there, but then it declined again [due to illness of former teacher]. Then the last three years. Today. What did I write? I like going to class. At school, it was horrible for me. I: What exactly was horrible? U: I was a lazybones [transl. lit. “lazy dog”]. I didn’t enjoy learning English. Ulf: 76-92 And even though he wrote in his language learning biography that recently and today he enjoys going to class, he still considers himself to be lazy when I, later in the interview, asked him to describe himself as an English learner: I: What else would be written about you as an English learner in that encyclopedia entry? U: Lazybones. [laughs] I: [laughs] U: Well yes. But it’s true. I admit this in all honesty. Ulf: 217-220 This quote also shows how learner Ulf can be positioned within the cate‐ gory “intended effort & investment” (sub-category of “Actual L2 Self ”). Within this category, participants seem to agree on how much effort they intend to invest in their language learning endeavour (from their perspective). Many of the learners do say that they are interested in the language and curious about English-speaking cultures (q.v. Dorothea and Anne below); they also cite this as an important characteristic of an ideal language learner. Ulf, for instance, is interested in the United States (he has travelled there before) and music in En‐ glish (e.g. country songs). Yet, when it comes to the actual effort or investment that young-old language learners have put into learning English in the past and intend to implement in future semesters, most of them view themselves diffe‐ rently. Namely, like Ulf in the quote above, they tend to view themselves as lazy or not very ambitious. 171 6.4 Results of the Interview Study A very distinct theme in Ulf ’s interview is that of “Independence & Self-de‐ termination” (a sub-category of “Actual L2 Self ”). This category refers to the differences between young learners and adult learners in terms of the power-re‐ lationship with the teacher and institution (connected to this is also the sub-ca‐ tegory “diachronic”). This reverse power-relationships is particularly well desc‐ ribed by Ulf, who states that in adult education as a “pupil” he is “on the better side”, because he is paying the teacher for the service of teaching him the lang‐ uage (see chapter 4: research context). Furthermore, he points out that he is too grown up and older than me and thus can expect a different way of communi‐ cating or a different type of classroom management. In contrast to the times at high school - when he would have hardly had any possibility, if a teacher mist‐ reated him or talked down to him - he is now in the position of a customer and does not need to “put up with that”: U: In adult education, you know, it’s just like that I as a student am on the better side. I: What do you mean by “better”? U: Well, I can tell you: “Well, Miriam, I don’t like what you are doing here.” I have no problem with that. I am the one who pays you in the end. There are really BIG differences. […] But as a pupil [at school] you may have to put up with that. Because then/ I am just too grown up for that. […] I mean maybe you have a certain ambition to say “Well, I want to teach him something.” But this has to be done in a different way. Ulf: 240-248 At the end of the quote, he acknowledges that, as a teacher, I probably do have the intention to push my learners and help them learn something. Nevertheless, he concludes, that this “urge” of the teacher needs to be communicated or done in another way in adult education contexts. In another part of the interview, Ulf again highlights how he - as an adult learner - is “old enough” to make his own choices. In this example, he chose not to do his homework as he now has different rights than he used to have at high school: You also noticed, whenever you gave us homework, I never did it. That’s just the way it was. I presumed to do that. I am old enough to do so. [laughs] Ulf: 54 In connection to this, Ulf told me that if the lesson is designed in a way that strongly goes against his expectations of an ideal English learning experience, 172 6. Words: The Interviews he would stop coming to class - as a last resort. This also included the intro‐ duction of regular homework and tests to which he was vehemently opposed: THEN I would stop coming. A test, and then also handing it in. Not with me. I wouldn’t come to class anymore. Guaranteed. If you had done THAT/ introduced that, I would have stopped coming. I had to do this too many times in my life. […] This shows you that I am someone who does not like to be forced into doing something. When so‐ meone is trying to make me do something I am not willing to do, then I won’t do it. I want to have fun doing it, a little bit. Otherwise, I would also/ and tests, for instance, or grammar exercises every time, I would go crazy. I wouldn’t come back. I wouldn’t come anymore. Ulf: 364-366 In contrast to studying at school and taking tests at earlier stages in his life, which have felt more like a duty to Ulf, he now wants to enjoy going to class without any additional pressure (e.g. homework or tests; Schmidt-Hertha 2014: 37 f., 66). This is also a in line with the sub-category “diachronic” (a sub-category of “internal frame of reference”): The central characteristic for coded passages within this sub-category is change and how it is perceived by young-old lang‐ uage learners throughout their language learning history. Ulf ’s coded passages illustrate the comparisons of how young-old learners themselves have changed as language learners, as well as how conditions for learning or the learning environment (i.e. the institution, language classes, material, teachers) have changed. With regard to the latter, learners in their interviews have compared the two institutions: school (where most of them first started to learn English) and the vhs centre. One common theme regarding this is that learning English at school is associated with obligation and pressure (in Ulf ’s quote above: talking about being forced into doing things and the pressure of tests that need to be handed in), and how for them this is the central differentiating characteristic between school and the vhs centre. Summarizing the above, Ulf shows a lack of purpose or weak connection to an L2 community (e.g. no regular travel) at the moment, a self-concept that includes being lazy and having no talent for languages, and finally a low wil‐ lingness to invest extra time into language learning outside the classroom (e.g. homework or preparing for a test): What made Ulf stay in this English class for almost three decades? The category “obligation” (sub-category of “external frame of reference”) can provide an answer to this: Then I was in such a low, of course. But I stayed for his sake [previous English teacher] and also for the GROUP’s sake, because we always said “We’ll stay together”. We always had too few participants. And sometimes we had to add another participant, 173 6.4 Results of the Interview Study who actually wasn’t there, and the others paid for that extra participant only so that the group could survive. That really is why I stayed. That was back then. In the time when it [his motivation] quite declined. Ulf: 46 Ulf ’s response is one example - apart from several other similar responses of other interviewees - of how the learning group itself creates a strong connection that helps overcome temporary motivational lows (e.g. due to a change in the teaching quality of his previously unwell teacher). It also illustrates the con‐ nection and need for every learner to come back in order to help the group to “survive” (see chapter 2: a minimum number of 6 or 8 participants is needed for a vhs centre course to continue each semester). In other words, this supports the view that the social aspects of attending language learning classes are par‐ ticularly important to older learners (see chapter 2.2; Murray 2011; Schmidt-Hertha 2014: 63 f.). 6.4.2.2 Medium L2-Vision Learners 6.4.2.2.1 Perfectionist and Realist: Wolfram Wolfram’s interview illustrates how the positions on the self of “own” and “other” (Taylor 2014; see chapter 3) can sometimes differ. He told me that he started with the conversation course in 1994 because he noticed how he was struggling to communicate freely at international congresses and how that made him feel ashamed. While he feels quite confident about grammar due to his more intense learning experiences with Latin and Greek at school, he sees his English vocabulary as his weakness. He repeatedly brings up the issue and how he con‐ siders himself to have “deficits” in this area: If I take a neutral look at my level of knowledge, I still see a problem with vocabulary. […] It has definitely improved. But it is an area in need of constant work, because one does not practice enough. Wolfram: 59-61 Looking at the meta-linguistic features of this quote (and further quotes in the following), he changes from the subjective “I” - talking from his individual per‐ spective - to a more general or objective “one” in this quote. It is an example that illustrates the change of perspective some interviewees perform in their interviews (main category “perspective”). Here, he is doing the change of per‐ spective, when he is talking about not doing or not having put in enough effort (by e.g. practicing vocabulary or, in a quote below, doing homework) - possibly to distance himself from what he says, i.e. shifting the responsibility for not practicing enough away from him. The change of perspective in this example 174 6. Words: The Interviews may also indicate remnants of an earlier ought-to L2 self, i.e. external socializers or significant others such as parents or his teacher at school, who once expected him to do his homework or practice more often outside the classroom (see chapter 3.2.3). Going back to vocabulary, it is interesting to see how Wolfram viewed himself critically in this area. In contrast to this, nearly all other interviewees from his class, who were asked about role-models, chose him due to his eloquence when speaking English. Wolfram himself named Barbara as a role-model because she knew many words he did not know. In other passages of his interview he told me he would like to expand his vocabulary, or often lacks vocabulary when trying to express his thoughts, which is why he comments on this with “here we go again”: W: I find Barbara impressive. I: What exactly do you find impressive about her? W: She knows so many words. [laughs] Here we go again. [laughs] Words that I don’t know. Wolfram: 148-150 He also mentions the issue of vocabulary deficits when talking about his ex‐ pectations about the course and his language learning experience. His answers in the following passage and the above quote (sections 59-61) indicate a medium L2-vision: he still sees a possibility of expanding his vocabulary skills - this is, however, happening at a slow(er) pace: W: But the great deficit was vocabulary, because one used to forget it again between travels and start again. That surely is something I expected from this course, that it improves. That I broaden and deepen my horizon. I: Has it met your expectations? W: Yes. But very discreetly. [laughs] I: [laughs] What do you mean? W: Or, very slowly. I notice this still today, and maybe that’s also because of ageing that I keep forgetting the meaning of words, or I cannot recall them the moment I need them. Of course, that is also a matter of focus. I assume. But that’s the way it is. The status quo. That’s why I am saying discreetly and very slowly, that the expansion of vocab, because one / and what adds to this is that one hasn’t done homework. Wolfram: 30-33 175 6.4 Results of the Interview Study He sees the reasons for his problems with vocabulary in English mainly as his age, and the lack of attention or focus in the evening class, but also as the lack of effort he has invested in such things as doing homework. Yet, another factor seems to impact him being able to express himself freely and finding the right words: perfectionism. Judging by Wolfram’s responses in his interview, it becomes clear that his perfectionism and, related to this, his inhibitions to speak freely and without a worry have different roots. Firstly, he connects this feeling and the feeling of shame or embarrassment with times at school and also a certain pressure to perform perfectly in his profession (e.g. when presenting at an international conference). Secondly, he describes himself as enjoying (he also uses the word “petty-minded” in his response) formulating and revising texts - both in German and English. Thirdly, as the following quote illustrates, he considers his cultural background to constitute a source for his perfectionism. Wolfram is aware of his perfectionism and explains this as being something typically German. He compares himself to Astrid, from Denmark, who has in his view far less inhibitions to speak and use the language: Undoubtedly. This is supposedly a German characteristic. […] Astrid speaks much more freely and with less inhibitions than myself. I always want it to be perfect, and preferably make no mistakes. But people aren’t like that. People from anywhere on the planet and who speak English. They just talk and try to make themselves under‐ stood. And if they are understood it is okay. Then they don’t care if they made mistakes or not. The main thing is they are understood. No, I just tend to be the way Germans are, and aim for perfection. And if I can’t do it perfectly, then I’d rather be quiet [transl. lit.: shut up]. I think one can probably say that I have (.) los/ partially lost THIS inhi‐ bition in the course, in this group, too. Not much, not completely, but at least partially. Still, I’m too German. [laughs] Wolfram: 279 The way Wolfram analyses his perceived lack of, and struggles with vocabulary throughout his interview also shows a high level of reflexivity (see chapter 3.4.1; q.v. Anne below). On a positive note, however, Wolfram mentions that he has observed a change as he - as well as other learners in his group - has lost some of his inhibitions in this conversation group over the years: But when - and I find this is a great advantage of our group - that there are no animosities. No jealousy. No higher expectations where someone sighs when another person can’t express himself properly. […] That was something very important in our group, because I had the feeling over those many years, that people - and also I myself 176 6. Words: The Interviews - definitely have lost their inhibitions with the time and always/ and have improved significantly when it came to speaking freely [transl. lit.: simply speak with courage]. Wolfram: 75 This quote again shows the dynamics of the external and internal frame of re‐ ference: In what ways are others around me developing? How do I position myself with regard to their performance or development? How have I developed over the years (“diachronic”)? - these are questions that Wolfram reflects on in this passage of his interview. Once more, the issue of being able to speaking freely - regardless of the number of mistakes - appears here as well. It is so‐ mething he seems to admire about other learners (even if their English skills or educational background are below his). Despite this critical self-evaluation, he also acknowledges that he has lost some of his inhibitions: because of the group or having become part of the group. Thus, in another part of the interview, I asked Wolfram - just like the other intervie‐ wees - to define his role in class (as though if the group were football team). Using the football theme, he described his role in class as someone who is a team-player and who “kicks-off” discussions in class by purposely offering opposing or unan‐ ticipated stances. By doing so, he is trying to challenge or stimulate others in class to participate in the debate and say the opposite. Wolfram also describes how, during these discussions, he is trying to look-up and incorporate new vocabulary not only for himself but also for others in the group so that they can profit as well: W: Well, I am definitely a team-player. Looking at the English group, I also consider myself to be a little bit like someone, who provides a kick-off [impulse, stimulus] for a conversation. I: How do you do that? W: Well a kick-off for a discussion, yes? Whenever a discussion starts, that I participate and try to develop points of views. I try to invite others in the group to join as well. Take part in the discussion maybe. To challenge to say the opposite, or/ well, I try to support a debate. In any case, I am also trying hard in terms of new vocabulary with its meaning to grasp or to look-up. Not only for myself, but also to make them clear to them, or name them, communitcate unknown words for the others to understand. So that they also profit from it. Yes? In that sense, I feel - maybe this sounds a little pretentious - I feel responsible to stimulate the others. Wolfram: 187-189 Wolfram has found a clear position within the group by being “responsible for stimulating the others” through taking a different stance in discussions as well as - a recurring topic - expanding his and the others’ vocabulary. Again, this illust‐ 177 6.4 Results of the Interview Study 9 “No, nothing has changed [after retiring] in ways I approach learning. Maybe I enjoy it a little bit more than I used to, because it is a welcome change in my everyday life, which reminds me of my working life, or of the foreign language that one wouldn’t use AT ALL otherwise. Hardly.” (Wolfram: 84) rates how the external frame (the social aspect, i.e. the other learners, the group) has shaped Wolfram’s L2 learner self-concept. At the end of his interview, he told me that the way he has taken on this role over time is also a sign of how he has become more comfortable of participating in class over the years and that he has lost his inhibitions. In view of upcoming semesters, he now wishes that the course “survives” and continues. This is after a short phase of fatigue and thinking about joining Ulf and Christine: two learners, who decided to discontinue after the current semester. He is now curious about how it will go on, and came to the conclusion that what the course has to offer goes “beyond the English language”: W: I will definitely try to keep the course alive. Because I profit beyond the English language and I see it as an opportunity the circle, which you have in life, that becomes smaller/ to overcome the limits of the shrinking circle - to looking a bit towards the left and to the right. I: How is it getting smaller? What do you mean? W: Because of ageing and leaving the workforce. Wolfram: 195-197 Here, Wolfram points to the social aspect of his learning experience and its incre‐ ased importance once he retired. As discussed in chapters 2 and 3, researchers claim that major life transitions such as moving from school to college play an impor‐ tant role in shaping a young person’s self-concept (Elder et al. 2003). Applied to the young-old, this major life transition is entering retirement and with it facing new challenges and possibilities 9 that can influence the self-concept. One such chal‐ lenge that Wolfram mentions here, is that of a shrinking social “circle”. To avoid increasing isolation, young-old learners need to find new ways to maintain and possibly extend social ties - Wolfram refers to this as “looking a bit towards the left and to the right”. Based on this, Wolfram seems to have re-conceptualized parts of his L2 self-concept by adding a different emphasis socializing by means of attending the English class (as opposed to the formerly stronger instrumental emphasis of such factors as work purposes). His previous statements illustrate the category of “connection” (main cate‐ gory: “external frame of reference”). The following statement shows a 178 6. Words: The Interviews stronger “obligation” characteristic (within the same main category as “connec‐ tion”): Well, it has to do with a sense of duty. It has to do with a responsibility towards the others. If everybody said: “Today, I don’t feel like going.” and submits to this, then only two people will sit there and that’s not okay. […] It is also a social event. And I am still interested in the English language. […] I (.) would have regretted it, if we had been too few members not being able to keep the course alive. Wolfram: 43-45 Because the English classes are a “social event” he feels obliged to attend regu‐ larly, as it is also a smaller learning group where the absence of learners is more noticeable when it comes to the dynamics of group discussions. Interestingly, in Wolfram’s responses here, but also in Adam’s and Ulf ’s interviews, the group is transformed into a breathing or living creature that needs to be kept “alive” through attendance and participation (“A communication class needs to be kept alive by a willingness to participate, by a commitment of the learners.”, Wolfram: 145). Adam, for example, mentioned another learner he often supported during class as she was struggling at times. This gave him a sense of purpose and one of the few remaining reasons to stay in class. Ulf remembered the group’s sen‐ timent (“We’ll stay together”) in earlier times of low participant numbers or low motivation. Furthermore, Wolfram, in another passage of his interview, empha‐ sizes the importance of extra-curricular events for the group’s energy, such as a trip to London and going to theatre plays in English once or twice a year: These surely are highlights. Because you appear as a group and then afterwards you discuss it and you also talk about how you noticed deficits and had difficulties un‐ derstanding. But it always creates this kind of impetus. An impetus to continue with it [learning English]. It pulls you out a little bit of this uniformity of this regular event on Mondays. Wolfram: 97 According to Wolfram, activities outside the classroom (he and other learners in this group also mention the regular visits to the pub after class) have streng‐ thened the connections within the group and gave participants new energy and an “impetus” to keep going. These examples show how learners become part of their learning group - grow together - and this way connect deeply. They define and re-define their role(s) within class. This, however, is a two-way street: the group also influences individual learners and, to some extent, their L2 self-concept (the case of Tanja, in the following, will illustrate this further). According to Wolfram and other 179 6.4 Results of the Interview Study interviewees, new members need to be able to make compromises or adapt their expectations to the group’s learning arrangement (e.g. pace, role of homework and tests, socializing) in order to become part of the group. Wolfram talks about how his group’s welcoming and accepting attitude towards new participants in the group - unless that person was trying to change the ways of doing things within the class: Not everyone fits into our group. That depends on his expectations towards the group. And on his level. And on his motives. So, when somebody joins our course, who expects to take something home every time, something to work on at home, and that is discussed the next time. I would regard this as counter-productive. Well, that may be exaggerated. But it would be difficult to introduce this in the group. Because they think differently. Especially because they have other everyday tasks to do […]. Then that new member must decide, whether this is enough for him with this group or not. Then he simply needs to leave again. But I think, that we generally greet them with open arms. That he will be accepted by us. As long as he doesn’t try to change our ways of doing things. Wolfram: 213 Interestingly, when talking about goals in this situation, Wolfram speaks for the whole group rather than only his own goals. It is not the only passage (and not the only interview) where it is formulated this way, or where he speaks in the name of the whole group in this way. I asked Wolfram what makes an ideal or successful language learner. Like Ulf, he also regards clear goals or a purpose as making successful English learning easier and giving learners greater determi‐ nation to reach their goals: Determination. Determination. Determination. (.) A strong determination to improve his English skills. Then it is easier for him if he has a goal in mind. If he wants to go to an English-speaking country and takes such a course. That would be such a goal. Then he would put real effort into it, because it is important for his future life. […] This type of goal does not apply to us. To our group. [lists a few examples from class] These are all vague goals. Not eminently important goals as in, for instance, such commercial language courses. Wolfram 131-133 When it comes to his attitude towards learning English and goals, or that of the others in his class, he evaluates these as less “urgent” and with it, makes a dis‐ tinction between vhs centre courses and “commercial language courses”. The contrast Wolfram creates between the vhs centre and other courses seems to involve the pace and type of learners (i.e. their motives and determination). This 180 6. Words: The Interviews links back to the discussion on the image of vhs centres in Germany in chapter 2: a very slow pace within courses for “bored housewives and retirees”, and “vhs centre brain: insult or compliment? ”. It seems, at this point, that Wolfram has internalized this public image in his young-old L2 self-concept to some extent. What this quote also implies, however, is that Wolfram, when describing what he considers an “ideal” or “successful” language learner, does not consider his English course at the vhs centre to be a place for this type of learner. How does this relate to his medium L2-vision? In the last and the following statements, he describes his goals as vague. At the same time, Wolfram shows awareness that it is also up to him and his ambition, how much effort he reali‐ stically puts into learning English (see sub-category “intended effort/ invest‐ ment”): It’s not like I would start thinking about not coming anymore, because I am not sa‐ tisfied with all of this anymore. But I would also return to the course, even if things are not going well - as an expression of a need for improvement. But it is a vague goal of mine. It is not - in my age - it is no goal. I definitely want to get so far, that in my working life or - that I don’t have anymore - or on holidays/ that I can speak freely with English and American people that I meet. This is, of course, a goal, but I don’t expect this from the course. Wolfram: 73 Here, Wolfram struggles to find a clear definition of his goal and of the likelihood of him ever reaching this goal: Firstly, attending the class is - to him - also an expression of not only coming when he is performing well, but also when he needs to improve on his skill. He then adds that it is a “vague” goal. Then he corrects this by saying that it is no goal - at his age. He goes on to say that, secondly, his goal is to be able to speak freely with native speakers on his travels. Yet, - on a more realistic note - this is something he does not expect from his current vhs centre English class, and that he would have to spend some time in an English-speaking country as he concludes in the following statement: Well I (.) of course I would like to see my vocabulary being expanded. Even though I know that I won’t achieve this ideal to be able to speak English in a totally relaxed manner anymore. […] I am aware that this goal cannot be achieved without a stay in that country and will never be achieved. And I can’t expect this from such a course. That goes without saying. But it is a possibility to broaden your horizon. To expand your English-speaking horizon and that’s why it makes sense and is recommendable. I/ magic, well. I’m afraid I’m too much of a realist that I could/ [laughs] Wolfram: 159 181 6.4 Results of the Interview Study In this passage, where I asked him what he would do if he had magical powers (interview question 13), his ideal L2 learner self became more defined. His ideal L2-vision of himself would be having the ability to speak freely to native spea‐ kers without feelings of inhibition due to his perfectionism. Despite his feeling that he can no longer achieve this ideal anymore, he still expects some expansion of his English skills or broadening of his English-speaking horizon. He further highlights this possibility of growth later in the interview, again pointing out that with regard to the upcoming semester(s) he needs to stay realistic: W: Firstly, I don’t know what it is going to be like with the new instructor. Secondly, I’m a realist, you know. [laughs] I: Meaning? [laughs] W: [laughs] What/ if I keep this up [learning English], maybe with a certain increase in terms of / a slow increase or expansion of my vocabulary, then I feel quite confident. Wolfram: 200-203 In contrast to the previous statements, this one indicates some hope - i.e. a medium L2-vision - in the way he still sees himself possessing the cognitive plasticity to (slowly) expand or maintain his vocabulary. 6.4.2.2.2 English as a Positive Side-Effect: Dorothea Dorothea is one of two learners presented here, who have only been in their English class for a short period of time (less than two years). Compared to Wolfram and Katja (see next section), Dorothea has only one passage coded “me‐ dium L2-vision”, which is her response to my question about her plans or goals for the next semesters: D: Well, I could definitely plan to study a little bit more. [laughs] Whether I can do it, I don’t know. I: What exactly would you study more? D: I would study vocabulary. I would expand my vocabulary. Dorothea: 148 Dorothea has a vague idea about the upcoming semesters. She mentions that she could study harder to expand her vocabulary, but remains sceptical about whether she will succeed. She is also formulating her response in a hypothetical and hesitant way (“I could”). Other coded passages in the rest of the interview, which will be discussed in the following, may point to why she is careful about formulating these plans to study harder (outside the classroom). 182 6. Words: The Interviews Dorothea joined the class together with her friend Anne (see below). Like Wolfram, she uses the expression “broadening horizon” when she talks about why she joined the class. On a general level, she aims to stay interested in a broad range of topics. This, however, puts her under time pressure: I think you need to have a broad base [be broad in scope]. As soon as you submit to narrow-minded views, or get your information from only one source - I find that critical. I always want to have a BROAD base. That’s why I never have time. Because there are too many things that interest me. Dorothea: 18 What do interviewees mean by using this expression - “broadening one’s ho‐ rizon” - more specifically regarding learning English as a foreign language? Upon asking for examples of ways in which the English class broadened her horizon, Dorothea seems to qualify her original statement: Well, of course, the expansion of my horizon with regard to the English course is, a little bit, a bit exaggerated. But, still, simply being together with other people. Ex‐ changing views. Learning something new. I believe that’s already it. Dorothea: 46 This emphasizes again the social aspect of joining a vhs centre course, and that learning English shifts more towards being a positive side-effect. Additionally, Dorothea adds to this that “in our age, it is particularly important, that you challenge your brain in every respect. This English course is one such way.” (Dorothea: 16) Yet, going back to her carefully formulated vision for future se‐ mesters, this broadening of her English-speaking horizon is best restricted to class time as she has several other obligations that broaden her horizon in other dimensions. Within Dorothea’s interview, we see how she changes her position within the category of “intended effort/ investment”: how much outside the classroom would I be able and would I want to study (e.g. homework)? How much pressure in forms of tests or homework would I be willing to accept, and to what extent would this actually be good for my learning progress? At the beginning of the interview she rejects the idea of studying outside the classroom to prepare for the next lesson by, for example, learning a specified amount of vocabulary: D: I think, if I WERE expected to learn three pages of new vocabulary by next time, that would be a reason for me not to go there anymore. Because I know for sure that I cannot do tha/ I wouldn’t be able to do that. Just like that, on the side. What’s the word for frog again? No, hedgehog. Hedge? What’s the English word? 183 6.4 Results of the Interview Study 10 A= Anne; both Anne and Dorothea were present during Dorothea’s interview, which was conducted at Anne’s home. I: Hedgehog. D: Hedgehog. Well, hedgehog. Dorothea: 48-50 Later in the interview she changed that sentiment to some degree and talked about how it may be a good idea to have a little bit of pressure in class. She hypothesized that this might stimulate participants to sit down at home and review the material. She spontaneously suggested a vocabulary test, but even‐ tually, based on her experiences with her other L2 (French), questioned the ef‐ fectiveness of this: D: A vocab test or something like that. I think one would maybe sit down at home and would/ but I need to add, of course, that you wouldn’t use such vocab frequently. That goes without saying. In the end that may not be effective. I don’t know, if I don’t need the frog - the hedgehog - if I never use the word hedgehog for a year. Hedge‐ cock. A: [corrects her] Hedgehog. 10 D: Hedgehog. That’s just the way it is. You forget. I know that from learning French, that I used to speak quite fluently. I need three, four days in the language, so that I recall the words. Dorothea: 154 Her statement also illustrates the notion of different L2-concept domains (Mercer 2011a; see synchronic category above) and how learning one L2 can influence the L2-concept of another. Dorothea’s learning experiences with French, which also included critical incidents in an L2-community, e.g. during her time as an au-pair in France as well as using it with a French-speaking employee in Spain, have taught her a little bit about her learning process (with regard to learning and recalling vocabulary in an L2). She applies this to her English L2 self-concept in that it helps her to estimate realistically the effort she would have to put into learning new words in English and how to do so more effectively, that is by going beyond a vocabulary test in class. We have seen with the previous cases of Adam, Ulf, and Wolfram that the lack of or diminished, direct contact with an L2-speaking community (e.g. by means of frequent holidays abroad or through work) seemed to limit their out‐ look on learning English at the vhs centre. As mentioned above, Dorothea has 184 6. Words: The Interviews learned French in the past (she spent two years as an au-pair in France) and feels more comfortable using it in contrast to English (see synchronic category above, and domains of self-concept in chapter 3). She uses English only to a limited extent whenever she is on holidays, or when she visits her son in Dubai. Ne‐ vertheless, experiences in the past (e.g. critical incidents in France) and her re‐ gular trips to Dubai have shaped her expectations regarding the vhs centre course as well as her preference for learning a foreign language. Dorothea fre‐ quently uses the terms “relaxed”, ”taking it easy”, and “learning by doing”. In the following statement, she talks about how she prefers to learn English in a playful way by speaking or discussing something in class, an opportunity she would not have at home: I am not a nerd. It’s not like I pick up the English book at home. I believe that you learn by speaking it, and you wouldn’t do that at home otherwise. […] You need to DO it by speaking. You pick up a word here and there. It’s not like I want to expand my English vocabulary like, I don’t know / just “learning by doing”. I’m taking it easy. Dorothea: 24 As mentioned above, Dorothea prefers a learning method where she is not aware of the learning process. This may be because she associates it with times at school, i.e. studying at home. The use of the word “nerd” in the above quote and the way she talks about her change with retirement in contrast to her behaviour in school and at work in other passages of her interview also point to this. In one such passage, where we talk about the change of power balance between teacher and learner (q.v. Ulf above), Anne (A) joins the conversation: A: Well we’re grown-ups. That is something different when you go to school as an adult than when it is children. D: That’s it. A: We don’t necessarily want respect for the teacher. [laughs] D: No, not at all. A: Well we do respect the teacher. But in a way that we submit to her? D: No, no, no. Dorothea: 112 Not only are the ideas of “homework”, “tests”, and “nerd”, but also having to submit to or respect (here referring to a type of fearful respect) the teacher, are things Dorothea does not want to encounter in her vhs centre class. She expects a different approach to learning at the vhs centre, one that matches her new 185 6.4 Results of the Interview Study situation of being retired (she claims at one point in the interview that she has become lazier, or more negligent with retirement) and one that takes account of the fact that English is one of many horizon-broadening free-time activities in her life. Thus, learning English as a positive side-effect appears to be a recurring theme within Dorothea’s interview. In her statement below she describes how she does not want to be aware of that “learning affair” while she is in class, but prefers for it to happen on the side (“incidentally”): Then it is always so interesting that you don’t think about it like “Do I have to do THIS now? ”. It’s fun. And because it is fun, you learn something new without having this learning affair on your mind. […] But it just happens incidentally. I find that very pleasant. Dorothea: 54 Similarly, Dorothea illustrates this using an animal metaphor when she talks about how she enjoys this indirect method used in the conversation class (“You stimulate us without us noticing it.”): We are like dogs or like animals. We sometimes get/ you sometimes drop something and watch what we make of it. That’s the way I perceived it. You stimulate us without us noticing it. That is a nice thing to do, by not being challenged so directly. But you do it in a playful way. You lure us. You throw it towards us and then we get going, depending on how we/ how interesting the topic was, or to what extent we are inte‐ rested in it. Then you correct this and that. I find this is a wonderful method. Dorothea: 108 She compares the group to a pack of dogs or animals jumping at whatever topic or activity the teacher throws in their midst or lures them with. This statement also highlights how she puts the focus on what is happening in class and the contents that are discussed at the centre of her and the other learners’ attention. In her interview, she vividly remembers heated debates about politics and how she enjoyed playing games in the last session. It appears that she also focuses on learning English in the moment rather than planning ahead, revising for upcoming sessions and setting goals or forming a future vision for herself using English on her travels. 6.4.2.3 Strong L2-Vision Learners 6.4.2.3.1 Something that “Really Exists” as an Incentive: Tanja Tanja is one of the few learners of the interview study with strong ties to an L2-community because her daughter lives with her family in the United States. 186 6. Words: The Interviews Tanja and her husband have visited them several times - sometimes for several months - over the last years. Her interview also has several coded passages indicating a strong L2-vision. In these passages, she mainly talks about how she has noticed a steady progress, and that she is confident that she keeps learning new things with every new semester in her English course at the vhs centre: I enjoy the course. Because of the other participants and also because I think that I profit from it. I could progress faster, if I studied harder, but I still notice a progress. It’s not like that [sighs] “I go there and it yields nothing.” It’s not like that. You always learn, every year you learn something new. Tanja: 265 I have/ for me it’s like that I tell myself “I learn a little something every time.” That’s enough for me. It DOESN’T HAVE TO BE/ I will never be perfect. I would need to be around English or American people. Tanja: 119 Similarly to Anne (see below), Tanja leaves room for the progress or growth of her language skills. She has noticed progress in the past which she also ascribes to the vhs centre course and is sure that she will keep profiting from it in the future. When Tanja talks about profiting from learning a foreign language and attending classes at the vhs centre, she not only refers to her travels abroad. Another common rationale among young-old learners in this study (see chapter 5) that Tanja sees as a benefit for her, is keeping the mind active. Mirroring what Markus and Nurius (1986) have described as a ‘feared self ’ (see chapter 3), some learners want to maintain their English skills through attending the English classes out of the feared possible self of loss or decline (for some learners this is connected to dementia). This applies to Tanja as well: Then I’m thinking that you hear a lot these days how problematic it is when you are getting older. I’m a little afraid of that. Because our daughter is so far away. What if you become confused in the head? As long as there is still a partner who can help you, that’s okay. But when you are alone it is harder. I think that with such things that you/ the English course and all, that it helps, TOO. That you stay fitter in your head. Tanja: 259 Tanja is worried about what would happen if she - one day without her partner and her daughter living in the United States - became “confused in the head”, i.e. developing dementia or Alzheimers disease. To her, learning English is one way to help prevent or work against this feared possible self, which is not ne‐ cessarily an L2-self in itself. 187 6.4 Results of the Interview Study As mentioned above, Tanja has been to the United States many times and spent much time in L2-communities. Thus, she has accumulated a relatively high number of experiences or critical incidents using her English skills. Passages containing these were coded “diachronic”. Tanja’s interview has 14 passages coded accordingly (the average number of passages with this category, is 9.7 and the highest amount within one interview were 17 passages coded “diachronic”). For many young-old learners in my interview study, critical incidents outside the classroom have given them an impetus for change. Often, these critical in‐ cidents abroad have been bad experiences for them, in which they realized that either the language skills they have acquired at school many years ago had not prepared them, or they have lost or forgotten much of their once much better English skills. In these instances, the consequence has been to attend vhs centre classes and brush up on their English Tanja has also experienced a critical incident abroad that gave her a new perspective on her language learning priorities in class. I asked her if there is anything she can do without in her English class (i.e. something she finds un‐ necessary) and further clarified, if she could do without, or with less, grammar in class. Her answer to this was connected to an incident when she was in the United States visiting her daughter and grandchildren. A misunderstanding between her and her daughter’s 80-year-old neighbour brought Tanja to the conclusion that grammar, or more specifically, knowing which tense to use can sometimes be crucial. She wanted to bake bread and needed rye flour. The neighbour found out for her where to buy it in the area and instead of replying “I will find it”, Tanja used the wrong tense (“I found it”) giving the neighbour the impression that her efforts had been in vain because she had already found and bought it: Because I realized that it doesn’t make sense sometimes. Different ideas in the right tense - You HAVE TO/ have I told you the story? I wanted to bake bread and couldn’t find rye flour. […] and then she [daughter’s neighbour] was a little upset in the sense of “I put in the effort to find out where to buy it, and then you have already been there? You have already found it.” I had said “I found it.”, but I actually wanted to say “I will find it.”, just like we would say “Don’t worry, I’m sure I’ll find it.” I’ve never forgotten this. I will, right? Because she was SO disappointed […]. Tanja: 227-229 At the end of this quote, Tanja stated that she will never forget this incident, as she upset the neighbour very much. She concluded from this experience that, despite once thinking that grammar is not that important, she cannot do without it when learning English if she wants to reduce the risk of such misunderstan‐ 188 6. Words: The Interviews dings in the future. The longer travels and stays in America also helped Tanja notice ups and downs in her language learning progress. She reported that her listening comprehension skills have improved considerably but that she still struggles to express her thoughts fluently due to a lack of vocabulary (i.e. her receptive vs. her productive skills). With regard to learning new vocabulary and critical incidents abroad, Tanja also reported small everyday learning opportunities that helped her expand her vocabulary and, due to a different contextualisation, helped her retain new words for longer. Playing with her grandchildren who live in the United States, for example, sometimes taught her new words or songs which she could still recall during the interview: Almost. (.) Almost, because they played baseball. […] and I always had to throw the ball to him and when he hit the ball we all cheered. And when he missed it, then he always said “almost”. [laughs] I also said the word when/ when I can’t recall it, then I immediately think about baseball and then: almost. Tanja: 239 This shows how she realized that learning a word in conjunction with an ex‐ perience or context made it easier for her to recall later by remembering the whole situation. Anne has reported a similar incident where she learned a new word in context, i.e. in a situation with a native speaker (see next section). Tanja also revealed extra efforts made to memorize new vocabulary outside the classroom when she was at home in Germany. She told me, for example, how she made lists of vocabulary she found hard to remember and stuck them to her kitchen cabinets, in her attempt to memorize them while cooking. As she admits in the following statement, however, she has struggled to keep up this routine after a longer holiday: Yes, but I don’t think I’m good enough. I am too lazy. I always plan to/ I don’t know. This year I thought: “You have to learn more vocabulary. You MUST.” Well, and I plan to do it, I do it for a little while, but then I neglect it again. That is/ I guess if I lived over there now with my daughter, then I would put more effort into it. And do it more intensively, so I can cope over there. But this now/ […] They [the vocabulary notes in the kitchen] have been hanging there for a while, because we were gone for four weeks. Ever since I am back, I looked at them well, occasionally. Maybe three, four times. But before that I really did it EVERY day when I was cooking. But I will go back to that. Tanja: 108-117 She views herself as excessively lazy because she has struggled to implement a long-term vocabulary learning routine for herself. However, she is not planning 189 6.4 Results of the Interview Study on giving up as she asserts at the end of the statement that she wants to resume her old routine of reviewing her vocabulary lists every day again. But she also learned another lesson after she returned from her longer stay in the United States and came back to discover that she had forgotten many of the words from her list, which she thought she had learned quite well: [laughs] And then I realized when I was in America just now, then I take a look at the vocabulary that is hanging there, and I was thinking “My goodness. You don’t re‐ member anything.” And that I had ALREADY forgotten some words, which I had memorized really well before. Because I didn’t need them at all. Tanja: 27 Like Dorothea (s.a.), she drew the conclusion that words are better retained long-term if she uses them or needs them regularly. Relevance to the real-world or everyday life also appears in other passages of Tanja’s interview. I asked her about “highlights” or particularly motivating moments she experienced in class. She reports how she once realized that a text about a choir in South America was based on true facts because she had already watched a report about it on TV before: You know, when I found out that “Gosh, in the English textbook that is/ these are all real stories. It’s not just made up. It REAlly exists.” These people and/ I saw them on TV. […] I already knew the story and this way I did not only learn English but I also learned about something that really exists. As if I had read it in the newspaper. Tanja: 95-99 In class, textbook material using real-life stories surprised and motivated Tanja. Her last remark seems to be related to her experiences abroad again and this explains why she remembers this moment as a highlight in class: Tanja told me at another point during the interview how she was sometimes trying to read the local newspaper when she was staying with her daughter in the States. She was actively seeking ways to improve her English skills while she was staying there. An intrinsic motivation or real-world purpose relevant to her constituted a central characteristic of her language learner self. Growing up and growing old may have contributed to this increased intrinsic interest in the world around her or broad interests (as Dorothea and Anne have reported as well): I notice in my age today, when I help my other grandchild with his homework, that I am much more interested than him. History. […] Then I thought “Gosh, you probably weren’t motivated back then and didn’t pay attention.” Tanja: 105 190 6. Words: The Interviews Her interests have shifted and re-focused as she grew older (see category “dia‐ chronic”). In terms of learning English, she now has a different connection and purpose in learning. Additionally, when she went to school she had no choice but to learn English as a subject, as she pointed out earlier in the interview. Now that she has a concrete purpose in learning the language, she is motivated to invest more time in learning it inand outside the classroom making use of opportunities to work on her skills (e.g. reading local US newspapers, keeping and revising vocabulary lists in her kitchen). However, the way she talks about her ways of learning outside the classroom seems to contradict her evaluation of the English course and her age group in general in the following statements: We are a little phlegmatic when it comes to learning I think. Well/ just think about how often we get side-tracked talking about our experiences. Disturbing the class [laughs] because we chat. Tanja: 85 Only someone who is not too ambitious will be happy in our class. Because that/ otherwise he wouldn’t feel comfortable in there. When someone has real ambitions, and wants to do a lot. We are too boring for that, way too slow. Because none of us HAVE TO learn it. We do it for the pleasure. For pleasure and partly to train your mind [transl. lit.: brain]. Tanja: 255 Like other interviewees (q.v. Wolfram), Tanja speaks not only for herself but for the whole learner group. She - like several other young-old learners in the in‐ terview study - sees the degree of intended effort within her English class as relatively low, with “pleasure” and socializing purposes as the main motives that draw this type of learner (i.e. 60to 80-year-olds as she postulates at the end of the interview; contrasting them with 20to 40-year-olds). But Tanja’s responses also reveal that this attitude of her English classmates seems to clash at times with her personal efforts and ambition. Tanja told me that she has contemplated leaving the course and instead joining a conversation course that is on a higher level (B2 instead of B1). She hoped that in that class participants would feel more confident in strictly speaking English only. Previous attempts to introduce this “English-only policy” in her current course had been rejected by less pro‐ ficient (and thus more insecure) individual learners, who said they would quit the course altogether: 191 6.4 Results of the Interview Study I already told you this with the other learner that she wanted to quit if we did things that way. Then ALL of us (.) said that we don’t want her to quit. That we/ that WE then say “Okay, then we won’t do it like that.” Tanja: 137 Yet these critical moments also showed the importance of keeping the group together (q.v. Ulf). As mentioned in the other young-old learner profiles above, in many cases the number of participants is important for a course to continue (especially in cases where a course has been running for several decades). Mo‐ reover, the social ties within the group are of great importance to the young-old learners who, for instance, seek to extend their otherwise shrinking social net‐ work outside the classroom (q.v. Wolfram). Tanja decided against a more chal‐ lenging course that would have possibly better matched her skills and goals. She thus decided this way in favour of the group: The group itself/ somehow we were a good match and it wasn’t necessarily the lear‐ ning but “Oh, you see each other, you meet others and you learn something along the way.” Sort of like that. What I hear from the others it seems to be the same for them. They are just looking forward to seeing each other again and that together you/ now we have been together so many years. Not all of them. […] The ones who have just joined are all new and they are a good match. Tanja: 57 This example illustrates how Tanja’s L2-self is shaped by a dynamic of her in‐ ternal aspirations, beliefs about herself and her age group, as well as external influences of other learners in her class. 6.4.2.3.2 Ambition as a blessing and a curse: Anne Anne joined the English class together with her friend Dorothea in 2013. She has travelled to several African countries, but planned to travel to Canada within the next years. This was one of the main motives - next to “train your mind a little” (see quote below) - for her taking up English. Her interview contains four passages coded “Strong L2-vision”. This sub-category of “L2-vision” was applied when learners - in formulating their responses, indicated the possibility of im‐ proving of their English skills. For instance, the use of “not yet” or “still”/ “yet” indicate that participants still see room for development, as the following two statements show: And I wanted to speak just ONE language WELL. (.) because I have already learnt several/ […] I am interested in many things, too. Also in the cultures of other countries. Anne: 35-39 192 6. Words: The Interviews I am not content. I still want to learn SO MUCH. […] Vocabulary. Grammar I think I can cope with. I mean, okay, I would like to be corrected, too. The most important thing is vocabulary - to expand it. Anne: 238-240 This sub-category was also applied to learners whose future plans and vision of themselves with regard to their L2 learner self were relatively defined or clear. In the following statement, Anne successfully creates a clearer picture of her future L2-self, in that she is able to describe what achieving her ideal L2-self would feel like: A: When I’m doing something like this today - like learning English now - then I want it to / I don’t want to progress so slowly [trans. lit.: drag myself about it]. I want to know MORE now and be able to do it better. That I progress faster. I: Have you achieved some of that already? A: A little bit, yes. Yes. I: Do you have an idea, what it will feel like once you have fully achieved this? A: That I can converse easily in all kinds of situations. Anne: 155-159 Also characteristic about the above statements by Anne is her ambition in her language learning endeavour. While most learners see themselves as lazy or not ambitious, a few learners evaluate themselves in the opposite direction. Anne is one of those who see themselves as ambitious. In her interview, she reports how she was ambitious earlier in her life and how this ambition also impacted how she raised her son. Nevertheless, she views this also as a disadvantage at times because she feels being overly ambitious makes her feel anxious at times and so she avoids coming to, or participating in, class: I am actually ambitious. But sometimes I am TOO ambitious. So that I prefer to withdraw thinking I am not good enough. Anne: 29 To exemplify the negative effects of her ambition in English classes she talks about a speaking task routine we performed at the beginning of most lessons: learners could choose from a range of smileys/ emoticons to indicate how they felt that day. They would then take turns and talk about why they chose the particular emoticon and how their day or last week had been. Anne described the mental struggle she used to experience while the others before her took turns. Every time, she wanted to express in English exactly how she felt and 193 6.4 Results of the Interview Study what she had in mind. Yet, on every occasion she could not attain her idealized phrasing, but rather formulated a shortened version of it due to a blockage caused by being overly ambitious: But it is also on my mind already when the others take turns. Then I’m thinking about what I’m going to say. I was always a little anxious: “Now TODAY you can do it. You can express it THE WAY you feel it and how you think it.” Then I shorten it because I am so tense. [laughs] Anne: 269 In chapter 3.4.1, I argued for using Dörnyei’s (2005, 2009) L2MSS model in con‐ junction with young-old language learners due to their relatively high intrinsic motivation and ability for self-reflective thinking (in contrast to younger lear‐ ners, who have presently been at the centre of the majority of research in this field using Dörnyei’s model; Kormos & Csizér 2008; McEwon et al. 2014). In that regard, Anne is a case in point: The two statements above show her high self-re‐ flexivity, demonstrating that Anne is aware of her strengths and weaknesses. She is able to analyse phases of weakness where she withdraws from a course for a couple of sessions and she can express in words how and why she is blocked when she has to speak in class. But Anne has also learned how to deal with problematic situations like these, as the following paragraphs will show. The strategies that highly self-reflective young-old learners like Anne have developed over their lives to deal with difficult situations are connected to the concept of “resilience” (see chapter 2). In the literature, resilience has been desc‐ ribed as a strategy for coping with negative experiences and thus, for maintai‐ ning stability or facilitating growth (Mlinac et al. 2011; Nimrod 2015). Some learners display that they have incorporated resilience into their language le‐ arner selves by using strategies to motivate themselves or “pick themselves up”, and in this way preserve the self if encountering difficulties in their language learning experience. Anne, for instance, reports how she has learned over the years to pull herself up, if she is feeling down or lacks self-confidence (e.g. as in the situation she describes above when she stays away from class - “phases of weakness” - because she does not feel good or confident enough): I do want it. These are somehow phases of weakness, where I/ my self-esteem is weakened. It has always been this way for me, that sometimes I let myself go and then at one point I say to myself “No, that’s it. It can’t go on like that.” I have that inside me, that I pull myself up. […] and then [laughs] I have the feeling I push myself off from the depths of a swimming pool and then I am back at the top. Funny, isn’t it? Anne: 83-89 194 6. Words: The Interviews Here, Anne also gives an example of how she encourages herself whenever her confidence in her English skills is down: “No, that’s enough. It can’t go on like this.” She compares the feeling in such a situation to being at the bottom of a swimming pool and using her inner-strength or resilience to push herself back up to the surface. The theme of resilience is also strongly represented in the responses of other interviewees regarding advice to new language learners and when describing characteristics of an ideal language learner. Other learners point out the importance of willpower and of not giving up quickly (e.g. trying to attend more than one semester, or participating in class despite it possibly being at the end of a long workday). Another example of Anne’s self-reflexivity is her becoming part of the group (“connection” and “diachronic” categories). She joined the group only a year earlier and reports incidents that made her finally feel accepted in the group. At the beginning, she felt insecure about her role in class, as her daughter-in-law (Astrid, G1TF02) has been in class for a few semesters and convinced her to join with her friend Dorothea. Thus, in addition to her insecurities about her English skills being good enough for the course, she was also wondering if she would be regarded by the other learners ‘only’ as the mother-in-law of Astrid. She talks about how this changed when, for example after almost a year, she enjoyed a game they played in the last lesson, and how she felt part of the group. Additi‐ onally, she remembers how another learner - Christine, who has been in class for many years - once encouraged her to be patient with herself and stay in class. In another situation, Christine asked her for a recipe and complimented her on something she baked. At the end of the quote, however, she reflects on the whole process of settling into the group and assumes that it is her low self-esteem that has also played into this: Just like last time: the game. It was so funny. Then I had the feeling, let’s say, to be accepted in the group. I didn’t have that feeling at the beginning. I thought that maybe they think that’s my daughter-in-law and that’s the mother-in-law, or whatever. I don’t feel very integrated then. […] Also when Christine asked me for the recipe at Astrid’s birthday. They were the best pastries. They were mine. And I thought: “Well, there you go, they discovered something good about you.” [laughs] It’s always the self-esteem that seems to influence me negatively. Anne: 93-95 Moreover, she noticed during her two semesters in this course that others are also not perfect, despite the fact that she is still worried about making too many mistakes. She has arrived at the conclusion that she has become stronger, that she has warmed up with the group, and that she is now accepted: 195 6.4 Results of the Interview Study Now I have become stronger [knocks on table]. Now, in the group I have/ I have con‐ nected [transl. lit.: warmed up] and I have noticed I’m accepted. Anne: 99 After listening to Dorothea’s interview (s. a.), Anne contrasts herself in her re‐ sponse to the question about her role or position in the group (using the hypo‐ thetical football team): A: I am more in the back. […] Maybe midfield. […] I have the feeling I jump into the gaps […] not at the front like Dorothea - this wouldn’t be me. I: Why not? A: No. That’s the problem again. It’s the self-confidence again. I don’t think that/ well, it’s also like that I RARELY sit in the first row. If any, it is the second row [laughs] even if all seats in the first row are vacant. Anne: 188-201 Clearly, her self-confidence is a recurring issue in how she conceptualizes her L2-self within the course and group. On the one hand, it is a good development that she has found a role in the group and that she sees herself as part of the “team”. On the other hand, she seemingly sees it as a problem that she is still inhibited in class. Unlike Dorothea, who runs and fights at the front, i.e. trying to (pro-)actively engage in discussions, Anne prefers to wait for situations where she is needed to “fill a gap”. Within the main category of “external frame of reference”, it was also relevant to what extent participants regard themselves as ahead or on an equal level as others, and if they have role-models as an inspiration inand outside the class‐ room. The identification can also occur in terms of having the same attitude to the learning goals, lifelong learning in general, and vhs centre classes. Anne, for example, transfers the comparison with others to the community or neighbour‐ hood she lives in: I don’t know. Maybe because/ I am so open. I am open for new things and curious. They only have their household chores and there are few/ but we are lucky now. Dorothea is here and in our corner in the neighbourhood are several, who are inte‐ rested in many different things. […] Somehow it’s quite common here [in this area] to train your mind a little. Anne: 69-77 In this instance, she identifies with others in her immediate environment, who are - just like her - interested in further education and training their minds. But she considers herself to be ahead of others her age, who “only have their house‐ 196 6. Words: The Interviews 11 For example, she writes text messages to her instructor in English whenever she can’t attend a lesson, or is late for class. She also told me that she sometimes writes messages to her nextdoor neighbour in English. hold” and are not necessarily in her neighbourhood, due to efforts to attend a variety of vhs centre courses. Thus, she regards her maintained curiosity about learning things (or lifelong learning in general) as something that puts her at an advantage or something she profits from. This self-perception also forms a foundation for her L2-learner self-concept. Finally, the statement below illustrates a benefit derived from her lifelong learning interests. In contrast to most of the other interviewees and in contrast to the deficit-model of ageing that has long prevailed in research and the public discourse (see chapter 2), Anne has noticed a positive change with retirement: But the longer I am now confronted with English, with this language, the more I notice I can retain. Besides, in my opinion, my memory has improved overall. When I was still working so much, I was overwhelmed. I was a lot/ things were forgotten very much to my surprise. Now, all of a sudden, it is better again. […] It was the stress, too. I had my sick husband. I had the practice, my mother, and things on my mind. There were MANY/ apparently my brain filtered the most important things and the rest was lost. Anne: 53-57 Unlike several other interviewees, Anne has noticed that her cognitive abilities have improved after retirement. According to her, high levels of stress in her work and private life had a negative impact on her brain. What, from the analysis above, can we conclude made Anne’s L2-vision a strong one? Going back to what Ulf and Wolfram postulated about what made an ideal language learner, Anne has a concrete purpose or goal, a future journey to Canada, i.e. a direct connection to an L2-community in the future. Moreover, she displays a high degree of ambition, which on the one hand gives her the motivation to use English outside the classroom, 11 and on the other often inhibits her within the L2-classroom. Anne seemingly conceptualizes her self - and with this, also her L2 self - as being highly ambitious (often too ambitious) and cu‐ rious about a variety of things in life. Her case shows how retirement has freed up the necessary mind space for her to live up to this part of her self in classes at her local vhs centre. 197 6.4 Results of the Interview Study 6.5 Summary and Outlook The above analysis of selected young-old learners, based on the indicated strength of their L2-vision codings and how they are positioned within other categories, has revealed common and differentiating characteristics. It has also pointed out potential connections between these characteristics and how they may be connected to their L2-vision. In other words, what gives young-old le‐ arners a weak, medium, or strong L2-vision? While not aiming for a clear-cut learner typology here, the data analysis and presentation of purposely selected cases above can illuminate recurring characteristics of strong, medium, and weak L2-vision learners. These in turn can serve as a guide for educators in the field of foreign language learning to help recognize and facilitate vision-pro‐ moting factors inand possibly outside the classroom (see chapter 8). The data analysis of the interviews has also shown that young-old language learners as a target group are particularly interesting for investigating the lang‐ uage learner self and, more specifically in this study, the aspects of Dörnyei’s L2 Motivational Self-System (L2MSS; see chapter 3.4.1). The ability for self-re‐ flection and a strong instrinsic motivation are considered to be a good founda‐ tion for investigating Dörnyei’s model and the language learner self in general (Kormos & Csizér 2008; McEwon 2014). Young-old participants such as, for example, Anne or Wolfram display high self-reflexivity when it came to shed‐ ding light on their language learner self. Furthermore, as pointed out in chapters 2 and 3, young-old learners are more intrinsically driven to attend vhs centre classes particularly entering retirement. This is in contrast to younger learners, who still constitute the majority of study participants on the L2 self and to whom instrumental reasons (e.g. work purposes) are still playing a major role among their language learning motives. These instrumental reasons, which are extrinsic and therefore less interna‐ lised, can be related back to the ought-to L2 self component of Dörnyei’s L2MSS, i.e. what a learner perceives to be expected from him or her with regard to learning an L2. Considering the data analysis in this chapter, however, it can be argued that even the language learner self-concept of the young-old might have an ought-to L2 self. Broadening one’s horizon and doing something for one’s mind (i.e. learning a language to prevent dementia and similar cognitive im‐ pairments associated with ageing) have appeared frequently in the interview data. Such responses of young-old interviewees mirror the “successful ageing” trend discussed in chapter 2 (Rowe & Kahn 1997; Daatland 2005). When asked to elaborate on what they mean by doing something for their memory or broad‐ ening their horizon, young-old interviewees often qualified their earlier claims 198 6. Words: The Interviews or added that this was simply something they had heard or read about. Thus, learning English or attending vhs centre classes in order to live up to the ex‐ pectations prevalent in industrialized societies regarding active or successful ageing (see chapter 2.1.3) can be considered an ought-to L2 self of many young-old learners (accordingly, the survey study has shown that “keeping my mind active” is the most popular reason for respondents to learn English at their local vhs centre; see chapter 5). How have young-old language learners conceptualized themselves in the in‐ terviews as language learners? A common theme seems to be the question of laziness and ambition, or, intended effort and investment. While some young-old learners described themselves as lazy, they did not seem to feel particularly ashamed or bad about this. In fact, a few rejected the idea of introducing tests or homework in class and this of being pressured in this way into learning. It is something they strongly associated with being at school again and with it, the feeling of dependence on surrounding grown-ups (teachers and parents). It is a feeling that they no longer can, or want to, identify with as grown-ups. In some of Ulf ’s responses, for instance, we can identify a feeling of being rebellious when not doing homework, because he does not have time throughout the day, but also because he has the right not to do it - without having to fear conse‐ quences in the form of bad marks or letters to his parents. As a customer, he now has more freedom and - if he wanted - more influence. But the role of school and past learning experiences in a young-old’s learning biography can also influence their L2 self differently. Anne, a learner with a strong L2-vision, has continuously engaged in lifelong learning activities at her local vhs centre over the course of her life. Supporting her claim that she is - similarly to Do‐ rothea - interested in a variety of things, she has attended Spanish and French courses as well as photography classes. In line with what Hertha-Schmidt (2014: 37-39) reports, the learning biography of an older learner influences their be‐ haviour and attitude to lifelong learning beyond retiring, as well. What has been another central finding of the data analysis, is the impact of the social sphere when learning a foreign language at a vhs centre. Be it the shrinking social circle due to the retirement or death of significant others, or the hoped-for broadening of one’s horizon is through socialising: the group connects learners regardless of their L2-vision. In the case of learners with a weak L2-vision presented here, the group is the last tie that has made them stay for so long despite a feeling of decline in their English skills or motivation. In cases such as that of Adam, this seems to include the teacher as well, as he - next to two other learners in the study - sees the change in teachers as another reason for leaving the class and discontinuing English learning. But for other 199 6.5 Summary and Outlook learners with a better outlook on their L2 learning progress, the connection to their group is an additional, nonetheless crucial, reason to return to the vhs centre every semester. The group not only connects, but as we have seen in the selected cases above, also influences the young-old learners’ L2 self-concept. Interviewees seem to agree that their respective English classes are not the right place for learners with high ambitions. Yet, for individual interviewees with high ambitions (such as Anne or Tanja), this means finding a compromise between personal ambitions and the preferences or interests of the others in the class. This once more reveals the dynamic between the internal and external frames of reference when it comes to forming an L2 self-concept (Mercer 2011a; see chapter 3). Connected to the latter, another defining characteristic of young-old learners with a strong L2-vision, i.e. Anne and Tanja, seems to be their willingness to spend time learning English outside the classroom, even if they do not always succeed in keeping up the habit. It is arguable, however, whether these learners have always been this diligent when it comes to learning (e.g. at school), or whether the connection to the L2-community gives them the necessary drive. In the case of Anne, the former seems to apply: she sees herself as always having been ambitious about learning new things in life or work in general. When it comes to Tanja, efforts she puts into learning English outside the classroom have been sparked by the increased relevance of English as a foreign language in her life, i.e. her family abroad, whereas she described herself as not greatly interested in subjects such as English or history when she was a student at school. When we take a closer look at the two learners represented above with stronger L2-vision than others, the connection to an L2-community or the conc‐ rete plan to visit a country where the target language is spoken are relevant factors. The subject matter itself is more or at least equally relevant when com‐ pared with other motives (such as active aging purposes and socializing), though not the most important reason for continuing to learn English. Having a more direct, and rather frequent, connection to an L2-community, or concrete travel plans appears to make the ideal L2-self realm more defined and necessary to attain (see chapter 3.4.1). In other words, for Anne and Tanja, learning English goes beyond being merely a positive side-effect, mitigating an isolated or cog‐ nitively impaired future self (feared self). Learners like Wolfram and Ulf, who would not necessarily fit into this category as much as Anne or Tanja, have pointed out that it is a purpose or a concrete goal like this that provides language learners with the necessary determination. Finally, it is worthwhile evaluating the value of qualitative inquiry in the context of this study and determining what the findings in this chapter mean 200 6. Words: The Interviews for the final merging of both quantitative and qualitative data strands in the next chapter. The qualitative content analysis, which was chosen for this study, has systematically reduced the vast amount of interview material. It helped create a category system (i.e. a network of themes relevant to the research interests) and link the interviewees’ responses to the theoretical concepts outlined in chapters 2 and 3. Going back to the statement by Richards at the beginning of this chapter, the data analysis of the interviews and the discussion of selected participants has shown how qualitative inquiry can “probe beliefs and experi‐ ences that might explain their [the learners’] responses.” (Richards 2009: 187) In this way, the qualitative data collection and analysis in this chapter have pro‐ vided additional, in-depth, and possibly alternative perspectives on young-old language learners in relation to the quantitative part of this research project (Duff 2012: 98). If and to what extent both data strands act as counterweights, which has been postulated at the onset of this chapter, will be explored in the next chapter. 201 6.5 Summary and Outlook 7. Numbers and Words: Merging QUAN and QUAL Strands After the analysis and interpretation of the individual data strands - quantitative (QUAN; survey) and qualitative (QUAL; interviews) - this chapter brings both strands together in a discussion to eventually form meta-inferences in the light of my research questions (see figure 7.1 below; Teddlie & Tashakkori 2008). As illustrated in chapter 4, bringing two strands together can result in both con‐ vergent and divergent findings - even if it is on the premise that theoretical constructs have been equally well operationalized in both research instruments (construct validity). It is the aim of this chapter to reveal those varying degrees of convergence and to discuss possible reasons and consequences with regard to a) the study design, and b) older foreign language learners and foreign lang‐ uage learning in general. In section 7.2, I will use the findings on the survey and the interviews outlined in previous chapters, as well as the merging of the two data strands in this chapter, to find answers to my research questions. Figure 7.1 MMR Research Design (Final Merging Stage highlighted in green) Creswell and Plano Clark (2007: 137) suggest that in order to merge different data strands, one can either transform one data strand (e.g. transforming textual data into numerical data), or compare both data strands without transforming them (by using e.g. a matrix or a discussion). For the purposes of this study, in which QUAN and QUAL components are weighed equally, the data strands will not be transformed but rather compared in a discussion. In this way, both strands are considered in their original numerical or textual nature bringing their res‐ pective strengths to the merging process. However, while data strands have not been transformed, the merging and comparison in section 7.1 below will take a different perspective on the results of both data strands. To ease the merging process, the findings of the Likert scale items of the quantitative data strand will focus on individual items in a descriptive form. This is an alternative to the 203 7. Numbers and Words: Merging QUAN and QUAL Strands multi-item scales that have been correlated with other factors in chapter 5. Ne‐ vertheless, the results of the multi-item and adjective scales (apart from the descriptive statistics presented in chapter 5) are still relevant for setting the focal points for the discussion below. A similar procedure has been chosen for the qualitative strand. The analysis in chapter 6 involved the creation of profiles of six learners that have been selected based on their L2-vision. Like the multi-item and adjective scales in the quantitative data strand, they will set focal points for the discussion below. Yet, the merging of the data strands will showcase a cross-section of all 21 young-old interviewees. It is hoped that the re-focusing from this different perspective can clearly exemplify the “friction” - i.e. con‐ vergent and divergent tendencies, as well as differing perspectives - between the two data strands (Knorr & Schramm 2016: 90 ff.; Creswell & Plano Clark 2007: 140 ff.; see chapter 4). In sum, central findings from both quantitative and qualitative data strands (see chapters 5 and 6) have provided stimuli for the merging procedure and guide the discussion below. In this way, both strands help to purposively select from the vast amount of data in both strands, i.e. aid the presentation sampling for this chapter’s discussion (Legutke 2016: 389). Concluding from this, the data discussed below have been selected on the basis of • overlaps of quantitative and qualitative data strands, • conspicuous responses in the quantitative data that call for further depth through qualitative findings, and • strongly emergent themes in the qualitative data that call for a confir‐ mation or disconfirmation in the quantitative data strand. These selection criteria also mirror the rationales for employing MMR discussed in chapter 4 ‘Completeness or Expanded Understanding’, ‘Offset’, and ‘Inference Quality’ (see chapter 4.1.2). Based on the above selection procedure, chapters 5 and 6 have yielded the following focal points for the merging of both data strands: • Ambition and Intended Effort • Resilience • Critical Incidents Abroad and L2-Vision • Confidence: Taking it Easy • Defying the Effects of Ageing • Now and Then: Ways of Learning a Language • Now and Then: Learning Context • Perceived Language Learning Progress • Language Learning and the Social Sphere 204 7. Numbers and Words: Merging QUAN and QUAL Strands 7.1 QUAN-QUAL Integration of Micro-Level Results 7.1.1 Ambition and Intended Effort QUAN 38 QUAL I: What else would be written about you as an English learner in that encyclopaedia entry? U: Lazybones. [ laughs ] Ulf: 217-218 I don’t study. I go to class. [ laughs ][ … ] I say this in all honesty. I don’t sit down and study. I don’t do that. Jennifer: 148-168 I am actually ambitious. But sometimes I am TOO ambitious. So that I prefer to withdraw thinking I am not good enough. Anne: 29 3 16 47 42 33 12 0 10 20 30 40 50 disagree 2 3 4 5 agree I'm ambitious when it comes to studying English. 0% 10% 20% 30% 40% 50% 1 2 3 4 5 content (1) ambitious (5) a (less than a year) a (1-5 years) a (more than 6 years) Table 7.1 Merging Results: Ambition and Intended Effort The two graphs in the quantitative data strand show that participants tend to view themselves as ambitious about learning English. In the first graph, most 205 7.1 QUAN-QUAL Integration of Micro-Level Results respondents agree with the Likert scale item “I’m ambitious when it comes to learning English.” A similar trend can be seen in the second graph, the adjective scale, with participants who have only joined their course recently sticking out on the “ambitious” side. In the responses of the interviews, however, young-old language learners are drawing a different picture. Except for a minority (q.v. Anne), they do not describe themselves as ambitious (anymore) with some even saying that they are lazy, or have become more negligent with entering retire‐ ment (q.v. Dorothea). Interviews show that most learners are interested in lear‐ ning the language, but they describe themselves as rather relaxed, or even lazy when it comes to actual studying or investment. This appears to apply particu‐ larly to investing time in learning English outside the classroom (q.v. Jennifer) as it seems to remind interviewees of being at school again (q.v. Ulf). Some learners in their interviews agreed that they could or should be more ambitious about their learning. Regarding ambition and intended effort, then, the two different data strands diverge on the question of ambition for language learning. One possible reason for this divergence could lie in factors such as research relationship between the researcher/ teacher and interviewees, on the one hand, and response bias (social desirability), on the other hand. Since survey participants remained anonymous, their responses could therefore be impacted by social desirability towards these questions. Interview participants, however, have known me as their teacher for a longer period of time and possibly felt comfortable in openly admitting that they perceive themselves as “lazy” or not sufficiently ambitious. The divergence in the findings could also be explained by the fact that questionnaire respondents have contextualised the question differently, i.e. that they feel ambitious about their language learning endeavour - particularly when they are in class, but they might agree with interviewees when it comes to the willingness to do homework. 206 7. Numbers and Words: Merging QUAN and QUAL Strands 7.1.2 Resilience QUAN QUAL [ on what makes an ideal English learner: ] Determination. That he doesn’t throw in the towel quickly. Anne: 151 Then I was sceptical: A conversation class in English? I have never done anything like this. And haven’t done anything in a while. I thought I wouldn’t be able to follow. But then I just gave it a try. The older I get, the braver I become. Jennifer: 8 [ giving advice to new learners: ] Definitely sticking to it and not giving up. Even if you sometimes think that you can’t always follow. Stick to it nevertheless and try. [ … ] Yes, I am more the type of person who is tenacious [ trans. lit.: bites herself through ] . Who sticks to it. Brigitte: 262-264 That was indeed bizarre [ transl. lit.: adventurous ] . We’ve had interesting teachers. When you look back at it. It was/ Well. [ laughs ] But we are still here. That’s important. I’m proud of that. We managed them [ previous teachers ] all. Brigitte: 278 17 39 60 16 15 2 0 10 20 30 40 50 60 70 disagree 2 3 4 5 agree Whenever I encounter problems while studying English, I give up quickly. Table 7.2 Merging Results: Resilience Interviewees report a growth in confidence and resilience as they have aged, which also impacts their language learning experience inand outside the class‐ room. Young-old learners report that they have acquired resilience as they ma‐ tured. These learners show resilience in difficult situations (e.g. when they travel or when they have struggled with a task or did not like a teacher or another learner) by employing positive self-assurance to pick themselves up and keep going. This is also a central piece of advice most interviewees would give to 207 7.1 QUAN-QUAL Integration of Micro-Level Results potential new learners. The quantitative results depicted in the graph above show the responses to the (negatively worded) Likert item “Whenever I en‐ counter problems while studying English, I give up quickly.”, which overlaps with the construct of resilience from the qualitative data strand. 78 percent of all respondents between 60 and 80 years of age are on the “agree”-side of the above graphical representation of all responses. Results of both data strands point in the same direction: in the survey, respondents clearly indicate that they do not give up quickly if they encounter problems in their language learning process. Related to this, the theme of “resilience” has emerged in the qualitative data as a sub-category of the actual L2-self. Adding depth to the survey data here, the interview study participants often readily connected their attitude of not giving up to their (psychological) ageing or maturing over the years. 208 7. Numbers and Words: Merging QUAN and QUAL Strands 7.1.3 Critical Incidents Abroad and L2-Vision QUAN QUAL If it is colloquial, yes, I can keep up here and there. But of course if it is getting a bit more concrete. I would have to say, no, then I would need my dictionary so-tospeak. Dorothea: 136 3 3 18 35 64 31 0 10 20 30 40 50 60 70 disagree 2 3 4 5 agree I can imagine a situation where I am using English with people from abroad. 2 2 30 49 42 29 0 10 20 30 40 50 60 disagree 2 3 4 5 agree I can imagine myself using English when travelling abroad and effectively communicating with the locals. Table 7.3 Merging Results: Critical Incidents Abroad and L2-Vision Maintaining one’s English skills and small steps towards more vocabulary or fluency are typical future L2-self descriptions of young-old interviewees. This way, most interview participants formulated their goals in a humble way (e.g. talking about a light far away at the end of a tunnel, or about slowly but surely expanding their knowledge or improving their skills). Other interview partici‐ pants formulated their goals in a rather disillusioned fashion, seeing hardly any chance of considerably improving by only attending the vhs centre classes. Since many of the interviewed learners have already travelled and used English ab‐ road, they are able to make predictions about their future chances of performing 209 7.1 QUAN-QUAL Integration of Micro-Level Results in the language based on their experiences. In the case of Dorothea’s quote above, she is confident that she can converse about everyday (colloquial) topics but would need a dictionary as soon as conversation became too specialized. Critical incidents abroad have served several interview study participants either as a warning to improve or refresh their English skills, or as a surprise/ encou‐ ragement when they were performing better abroad than they would have ex‐ pected (based on bad experiences at school or earlier English classes). The survey draws an optimistic vision or ideal L2-self with regard to using their English skills abroad in both Likert scale items represented above. Most 60to 80-year-old survey participants agreed or somewhat agreed that they can pic‐ ture themselves speaking English outside the classroom. The number of res‐ pondents who (absolutely) disagree with the statements is particularly low for both items. Even though the vision-category in the interviews is dominated more strongly by a realistic/ maintaining vision of the ideal L2-selves, this does not necessarily indicate strong divergence of both data strands but rather an instance of expanded understanding (see chapter 4). The diachronic category, for instance, depicts young-old learners’ multifarious critical incidents abroad and in language classes over the years. Based on these experiences, it can be postulated that they, on the one hand, are aware of their capabilities when using English abroad in the near future (related to this: L2 self-efficacy - see chapter 3). On the other hand, their experiences have taught them to form more realistic goals for their situation, such as maintaining the already acquired skills that enable them to communicate successfully abroad, or slowing down the pace and taking off the pressure with the transition to retirement. 210 7. Numbers and Words: Merging QUAN and QUAL Strands 7.1.4 Confidence: Taking it Easy QUAN QUAL Then I dare to speak. Why not? If it is wrong, then it’s wrong. I don’t care. I have noticed that they couldn’t do it better than me. Ulf: 142 Because I am just older and I actually don’t care what other people think about me. I am what I am. When I say it that way, then I SAy it that way, yes? Barbara: 144 I am not inhibited anymore to speak. Even if it is wrong sometimes. Lena: 196 0% 10% 20% 30% 40% 50% 60% 70% 1 2 3 4 5 inhibited (1) relaxed (5) a (less than a year) a (1-5 years) a (more than 6 years) 3 3 18 21 61 43 0 10 20 30 40 50 60 70 disagree 2 3 4 5 agree I approach learning English in a more relaxed way than in the past/ at school. Table 7.4 Merging Results: Confidence: Taking it Easy The first graph shows the adjective scale from the survey contrasting whether young-old learners felt either inhibited or relaxed in their English class. The comparison - despite being a cross-sectional snapshot rather than a longitudinal study - indicates that learners who have spent more than 6 years in their English class at the vhs centre are more relaxed than other learners, who have spent fewer semesters learning English there. Similarly, the Likert scale item puts this 211 7.1 QUAN-QUAL Integration of Micro-Level Results statement into a temporal contrast: the majority of learners agreed or absolutely agreed with the statement that they have become more relaxed about learning English over time (i.e. in contrast to school or earlier English classes). The coded interview passages show a frequent use of the words “ambitious”, “relaxed”, and “lazy” within this category. Furthermore, the sample statements above re‐ volve around having become more confident to speak freely over the years or with age. The interviewees above are no longer worried anymore about making mistakes in class or abroad. With regard to their expectations towards an “ideal” language class at the vhs centre, a relaxed atmosphere and approach to learning are important to many interviewees (e.g. no homework or tests, a comparatively slow pace; q.v. Ulf). Their accounts show that they do not take failures to heart as much anymore and tend to approach critical situations with more self-con‐ fidence and a more relaxed mindset. The survey and the interviews clearly point to the increasingly relaxed and more confident approach to language learning of young-old study participants, particularly if they - as the survey indicates on the adjective scale - have spent a longer time at the vhs centre. This attitude shapes their expectations of what English classes at the vhs centre should ideally be like. 212 7. Numbers and Words: Merging QUAN and QUAL Strands 7.1.5 Defying the Effects of Ageing QUAN QUAL C: I think it is positive to do something for your grey matter. Maybe I won’t get rusty so quickly. I: Have you noticed that it helps with your grey matter? C: No. [ laughs ] No, actually not. It’s something you read everywhere, you know? Christine: 152-154 Attention. You do need to focus in the lesson. And it’s good for the brain. [ laughs ] in that age. Lena: 240 You can do something again st dementia. Normal dementia [ in contrast to Alzheimer ] . So, that you can delay it as much as possible. You need to keep your head together. The mind needs to keep working/ be active. The brain needs to work constantly. For me, a foreign language is part of that. Barbara: 309f Apart from that I also do something for my brain. If find that’s a good thing when you learn languages. It’s supposed to be good. Diana: 16f None of us HAVE to learn it by all means. We all do it for pleasure. For pleasure or part of it is also that you train your brain. Tanja: 255f 145 138 95 93 68 62 60 46 31 28 19 7 0 20 40 60 80 100 120 140 160 Memory Travel Learning together with others I like the language Offer appealed to me Good experiences Enjoyment No other options in the area Friend/ partner Recommendation (Grand)child Work Motives for Learning English (60-80 years of age; n= 155) 3 7 15 47 59 23 0 10 20 30 40 50 60 70 disagree 2 3 4 5 agree I feel mentally more active and fitter because of studying English. Table 7.5 Merging Results: Defying the Effects of Ageing 213 7.1 QUAN-QUAL Integration of Micro-Level Results A common motive for taking up language learning is the interviewees’ assump‐ tion of preventing age-related, cognitive illnesses (e.g. dementia; q.v. Adam in chapter 6) or “broadening one’s horizon” (see for example Dorothea in chapter 6). Apparently, the active ageing trend (see chapter 2) has reached young-old language learners because learners from all three courses mention that they believe language learning has positive effects on their brain. This is in line with survey results in which the motive of “memory” was the top reason for learning English at the vhs centre (145 out of 155 respondents; see chapter 5.3.2 and first graph above). In the interviews, however, some learners (e.g. Christine) who mention this as a motive or rationale for their language learning, qualify this claim. Connected to this, in the responses to the Likert scale item above, most young-old respondents within the quantitative strands have noted the positive effects of their language learning endeavour. In sum, respondents in the quan‐ titative and qualitative strands seem to be aware or show an interest in the benefits of language learning apart from communication: keeping their mind active in old age and in this way defying the effects of ageing. It becomes clear through the analysis of the qualitative strands, however, that the realization of this motive and the hoped-for positive effects are not readily apparent to all young-old learners - unlike the trend of the above Likert scale. Even though some interviewees mention that they also hope to train their memory and keep their mind active, they do not perceive having improved much or at all, when reflecting on their actual situation. Again, the interviews present an offset or expanded understanding of a phenomenon in the quantitative data. The results show that young-old language learners overall have strong intentions and be‐ liefs with regard to such age-defying language learning benefits. 214 7. Numbers and Words: Merging QUAN and QUAL Strands 7.1.6 Now and Then: Ways of Learning a Language QUAN QUAL Back then it was real intense studying. Reading through vocab, covering up one side and several times “Well? ” until you had the word in your head. Now, I must say I can retain it better. In certain situations/ [ … ] It is TOTALLY different now. But when I look it up in the book then I do it as I used to do it back then, that I read through it like that. But it is much easier for me to retain it through experiences [ … ] . Anne: 137 2 9 22 43 62 14 0 10 20 30 40 50 60 70 disagree 2 3 4 5 agree The way I study English has changed over time. Table 7.6 Merging Results: Now and Then - Ways of Learning a Language Participants were asked how they approached learning English (inand outside the classroom), and in what ways they have changed their ways of learning or their language learning strategies since their time at school or in former vhs centre classes. Here, most survey respondents agreed with this statement. Most interviewees, however, could not think of anything they have changed in their learning strategies and some said they did not study outside the vhs course at all. Some described how they would try to learn new vocabulary by writing it down in class and - if they took the time - read it at home again (e.g. cover up the translation and test themselves as they used to do at school; see quote by Anne above). As discussed in chapter 5, in this comparison quantitative and qualitative data sets seem to diverge. A possible explanation could be the situ‐ ation of completing a survey vs. responding in an interview: while survey par‐ ticipants may have contextualized the question on their own terms and possibly felt - without being asked to give examples and thus not further reflecting on it - that this applies to them in general terms. Interview participants, however, when asked about if, and more importantly, how their ways of learning have changed, may come to a different conclusion after reflecting on this issue. 215 7.1 QUAN-QUAL Integration of Micro-Level Results 7.1.7 Now and Then: Learning Context 7.1.7 Now and Then: Learning Context QUAN QUAL I always liked learning the language. But during our time we lacked the opportunities to use it. And we didn’t speak so much in class then. Birgit: 64 We speak a lot more. Just/ during our time back then you didn’t have just recordings and all that and then LISTENing. I was missing that actually. It wasn’t like that when we learned English. [ … ] We wrote many summaries of some texts or translations. I must say that when I open up a book. Translations: that I can still do quite well. I hardly need a dictionary. Birgit: 89-94 You didn’t go into it so intensively. Also the English back then wasn’t so, how I can I put it? So up-to-date, so fundamentally important as it is today. You just did it. Because you had to. Wolfram: 23 26 34 40 17 16 3 0 10 20 30 40 50 disagree 2 3 4 5 agree The way English is taught at the vhs centre is comparable to the English classes I had at school. Table 7.7 Merging Results: Now and Then - Learning Context This example of comparing qualitative and quantitative data, exemplifies how the interviews can give more depth or detail to trends in the survey. The survey indicates that the language learning experiences young-old learners had at school (if they actually had the opportunity to learn it at school, since some of them only started with English later in life) differs from the current and more recent language learning experience at the vhs centre. The survey does not tell if this is perceived as a good or bad thing, nor does it provide information in what ways the two differ. The interviewees also compare their previous EFL learning experiences at school with the ones at the vhs centre and - in agreement with the survey study - note that things have changed over the last decades. The main differences that come to bear in the interviews are in terms of teaching methods (usually teacher-, grammarand textbook-centred; pressure to study for tests and do homework; focus on written language) and in terms of roles in 216 7. Numbers and Words: Merging QUAN and QUAL Strands class, e.g. the teacher-student relationships. Regarding the latter, some learners describe how they - as adults (in this case also several years older than their teacher) and as customers of the vhs centre - are now in a position of power. Interviewees talk about the differences between their learning and their learner role at school in contrast to their role at the vhs centre. Young-old learners often refer to the language learning experience at school as monotonous (e.g. an exc‐ lusive close adherence to a single textbook and no use of additional material, such as short stories or poems; focus on grammar, homework and tests; and a different teacher-student relationship). However, they mostly talk about ex‐ ternal ways of learning and not how they themselves have learned at school or in the past. In their interviews, participants repeatedly pointed out that they - as adult learners - are now in a different position as learners (see for example Ulf and Dorothea in chapter 6). It is more the role of a customer and no longer the role of a pupil at school who is pressured by passing tests or doing home‐ work. Young-old learners in the interviews show that they enjoy the freedom to approach language learning as they please, i.e. with less pressure and at their own pace, or to have more involvement in the decision-making processes in their English classes. 217 7.1 QUAN-QUAL Integration of Micro-Level Results 7.1.8 Perceived Language Learning Progress QUAN QUAL His wife is American. I can communicate with her without any problems. I understand pretty much everything - it wasn’t like that before. This way, I just notice that I have made incredible progress. Christine: 96f A: The curve is not correct here. I went down here again. I: Okay, you wrote here that you always forget a little or forget more. A: Yes. More every day. It is getting worse. Adam: 82-84 0 4 16 51 65 17 0 10 20 30 40 50 60 70 disagree 2 3 4 5 agree I feel that I'm making progress with my English. Table 7.8 Merging Results: Perceived Language Learning Progress Learners in the interviews appear to agree that, without the course, their English skills would be a lot worse today. A few have noticed considerable progress over the years - despite it slowing down or - in the case of Adam - even declining. Apart from Adam’s learner biography task, all other biographies of interviewees display a graph that has kept increasing or is at a high point at the moment of the interview (see appendix for examples). Generally, interviewees noticed their progress in particular when travelling abroad or when looking back at an earlier version of themselves just starting in the course with the person they are now (and how much more comfortable they feel to speak in class). The majority of interviewees can be categorised as having a realistic vision of their language learning prospects, i.e. that their main goal is to maintain their English skills with the help of the course. They see potential for growth or expansion of their skills. This is, however, only at a much slower pace compared to when they were younger. This is also reflected in the responses of the survey: most learners agreed with the statement that they feel that they have progressed. Again, quantitative and qualitative data strands show a similar trend, i.e. they converge 218 7. Numbers and Words: Merging QUAN and QUAL Strands in that learners have noticed progress in their language learning. The strong indication of progress in the quantitative data, however, does not give informa‐ tion about the intensity and the time-frame of their perceived progress. In the qualitative data we find agreement with the statement, as many interview study participants report how, for example, the vhs centre English class has given them more confidence and fluency. It also, however, shows that interview partici‐ pants’ experience slowed down, or very slow, progress as they grow older. This is also supported by the fact that many responses from interviewees were ca‐ tegorized as realistic, i.e. their focus is or has shifted to a maintenance of their language skills. 219 7.1 QUAN-QUAL Integration of Micro-Level Results 7.1.9 Language Learning and the Social Sphere QUAN QUAL It is fun to be together. Maria: 41 I: What motivated you to stick to it for so long? Or stay in this course for so long? C: Because of the group. [ … ] We always went for a beer afterwards. Back then I was still very house-bound. I had my husband’s grandmother and two small children. My husband was often away. To me, that [ English class ] was just getting out. Just like other people/ I never had the urge to do any certificates. Just like other people would go out bowling, I went to my English class. Christine: 7-10 The atmosphere is very, very important to me. Today, I say to myself: I don’t need to do this to myself. If that’s a group/ well, I was in an Italian course once. There were some “know-it-alls” in it. Jennifer: 74 But if that was a group, for example, that would find this [ homework and tes ts ] important, and I would be the only one who wouldn’t want to do this. That would be a reason for me to quit. And therefore it was always really great in the group. They were all like me. Brigitte: 80 145 138 95 93 68 62 60 46 31 28 19 7 0 20 40 60 80 100 120 140 160 Memory Travel Learning together with others I like the language Offer appealed to me Good experiences Enjoyment No other options in the area Friend/ partner Recommendation (Grand)child Work Motives for Learning English at vhs Centres (60-80 years of age; n= 155) 49 49 39 6 3 0 10 20 30 40 50 60 cooperative independent cooperative-independent 60+ Table 7.9 Merging Results: Language Learning and the Social Sphere 220 7. Numbers and Words: Merging QUAN and QUAL Strands The social sphere, which has turned out to be a central point of reference in the qualitative data analysis, has not been explicitly addressed with Likert scale items in the survey. Other parts of the survey, however, give an idea of how respondents view the social aspects of language learning at a vhs centre. Firstly, “learning together with others” is the third highest motive in the survey (95 out of 155 respondents chose this as a reason for learning English at a vhs centre). This resonates with Maria’s statement above: “It is fun to be together.” Secondly, the adjective scale cooperative-independent gives an insight into how young-old survey participants view themselves in their English class. The re‐ sults indicate that the majority of young-old learners consider themselves to be more cooperative than independent (or to value being cooperative more highly). These results appear to be in line with the qualitative data strand. Nearly all interviewees mention the group as an important reason for staying in an English class for a longer period of time (for some this means 30+ years). The atmosphere that is created through the teacher and other group members impacts individual young-old learners, and individual young-old learners adapt and sometimes in‐ fluence the group dynamic (q.v. quote by Jennifer above; see also the cases of Wolfram, Anne or Tanja in chapter 6). In the quoted statements above, for ins‐ tance, Brigitte highlights why she and the group are a good match: her attitude to learning English mainly in class without additional pressure to prepare homework or study for tests, is something she shared with other members in the group. The case of Brigitte, above, and the case of Tanja (see chapter 6) exemplify the notion of being cooperative in the survey results above. Becoming a team and, in critical situations, adapting personal preferences to that of the whole group in order to “survive” show how interviewees define being coope‐ rative rather than independent, i.e. acting without taking the interests of others into account. 7.2 Discussion in View of Research Questions As mentioned at the beginning of the chapter, this section will discuss implica‐ tions of the data analyses in the previous chapters 5 and 6, as well as the results of the above merging of quantitative and qualitative data strands. The following will therefore be guided by the research questions: How do the young-old conceptualize themselves as foreign language learners? a. How do they perceive their past, present, and future L2-self/ selves? 221 7.2 Discussion in View of Research Questions b. Which factors (e.g. age-related: retirement) influence their language learner self-concept? c. In what ways do survey and interview study converge? The last research question (c.), is not only different in nature to the other ques‐ tions as it is a methodological research question, but it also seems to imply that the data strands do converge. As discussed in chapter 4 and detailed in the dis‐ cussion above, this could never be presumed in a mixed method study like this, and that it is in fact the divergent findings (or the “friction” mentioned in chapter 4) that can point to extraordinary findings. In the divergent findings above, for instance, it often appears to be a differing depth of self-reflection in the survey and interview situations that may have led to opposing views on a particular theme. Interviews - which have been conducted between learner and resear‐ cher/ teacher - allowed for clarification, elaboration, and “on second thought”-moments. In convergent findings, the above merging procedure has exemplified that using survey and interview results in conjunction can indeed provide an expanded understanding of complex subject matters, as is the case with the L2 self-concept of young-old learners. The inferential statistics in chapter 5, showed that older language learners (71 years of age and older) had better past L2 self-scores than younger respon‐ dents in the survey. Furthermore, the analyses of both actual and ideal L2 self multi-item scales show a proportional increase with the time spent at a vhs centre. Interestingly, other factors such as retirement or gender did not yield significant differences in the survey. The qualitative data analysis revealed the dynamics between past, present, and ideal L2 selves. As discussed with regard to the formation of possible selves in chapter 3, for instance, past language lear‐ ning experience and the actual L2 self-concept also feed into young-old language learners’ future self/ selves. The interview study has provided an insight into how past language learning experiences and critical incidents have provided a solid foundation for building self-confidence and resilience when using the language in class or abroad. Their past language learning experiences also feed into lowered expectations, or, an already approximated actual and ideal L2-self, which is in line with Ryff ’s (1991) study discussed in chapter 3. Most learners envision their future learning progress as increasingly smaller steps (if they expect any further learning progress at all), or as a maintenance of their current level. In their opinion, the vhs centre classes serve this goal well. Survey study participants display a similar trend by indicating an increased feeling of con‐ tentment over ambition (see chapter 5). Regarding the dynamics between past and actual L2 self, survey respondents indicated that they see a difference between the ways English is taught at the 222 7. Numbers and Words: Merging QUAN and QUAL Strands vhs centre and the ways it has been taught e.g. at school or in English classes they attended when younger. They also reported a change in the ways they approached learning English today compared to earlier attempts, i.e. they feel more relaxed and have changed their ways of learning English (e.g. in terms of learning strategies). The interviews once more give more details and depth to these responses in the survey. In the interviews, we can see how their past le‐ arner selves (being a student at school) sometimes clash with the actual L2 self of a grownup language learner. As mentioned above, young-old interviewees prefer to distance themselves from their past student L2 self by showing a sense of self-determination when it comes to deciding on their intended effort in re‐ gard to their language learning endeavour (e.g. homework, tests, slower pro‐ gression/ pace, and thus less pressure). Nevertheless, what also plays into these expectations of “slowed down” learning appears to be the public image of vhs centres in Germany (see chapter 4.3). The self-determination displayed in the interviews (q.v. the case of Ulf, Dorothea, and Tanja), however, needs to be ba‐ lanced with the social sphere of the classroom, i.e. there is often a need for compromise with other learners in class, who may differ in their intended effort and expectations of the vhs centre class. Related to the second research question, which factors influence the L2 selves of young-old language learners has also been of interest in this study. Apart from gender and the learning institution vhs centre, age-related factors across the different ageing dimensions discussed in chapter 2 (sociological, psycholo‐ gical, and chronological) were also relevant. The survey did not support the claim that the factor of retirement (which can be connected to the sociological dimension of ageing, see chapter 2) has a negative impact on L2 self-concept and its temporal facets. Except for Anne, who experienced an increase in her cognitive capabilities, most interviewees saw a decline in their intended effort after they retired, i.e. they have become more relaxed (if formulated positively) or negligent when it comes to learning English. The psychological dimension of ageing (i.e. the growth of resilience and self-confidence, as well as the incre‐ ased sense of self-determination discussed above) has also proven to be a rele‐ vant factor impacting the young-old L2 self; through their general maturation process over the life course and more specifically through critical incidents when using the target language, many young-old survey respondents and in‐ terviewees report that they have grown in resilience and self-confidence (con‐ nected to this is also the increased self-determination discussed above). Critical incidents abroad and anticipated travels in the future especially appear to im‐ prove their language learning vision. 223 7.2 Discussion in View of Research Questions Furthermore, the group, or social sphere, is another crucial factor that impacts the L2 self of young-old learners. While the survey only covered a few aspects of this factor, the interviews revealed the importance of a connection to the group, for example, for long-term members. The learning experience in the group - how young-old learners position themselves within the group, the con‐ nection or feeling of obligation, becoming part of the group, finding compro‐ mises in learning in order to “survive” - has a considerable impact on their language learner self. Finally, how can these results be connected to the L2 motivational Self-System (L2MSS; Dörnyei 2005, 2009) presented in chapter 3? The L2MSS consists of the three components of ideal L2 self, ought-to L2 self, and L2 learning experience. The first two components - ideal and ought-to L2 self - are future self-guides, and in this way, are related to L2-vision within this study. The discussion above and in the last two chapters showed that ideal L2 self is particularly influenced by past and actual L2 self, which encompass, for example, critical incidents using the language in the past, perception of the vhs centre courses or institution, and self-perception with regard to intended effort or ambition. The future self-guide ought-to self, however, has not been explicitly investigated in this study nor has it been addressed in detail in Dörnyei’s work. Dörnyei (2009) has focused more strongly on the motivational potential of ideal L2 self rather than ought-to L2 self. On ought-to L2 self, he writes that it “concerns the attributes that one be‐ lieves one ought to possess to meet expectations and to avoid possible negative outcomes” (2009: 29), and that it thus encompasses more extrinsic motives for language learning. From the data analysis and merging we can see that the above theme “defying the effects of ageing” reflects the understanding of ought-to L2 self and to what degree it has become a component in the young-old’s learner self-concept. The statement by Christine in section 7.1.5, for instance, exempli‐ fies how some young-old feel the need to keep learning after retirement to re‐ duce the risks of cognitive impairments (decreasing memory or dementia) - i.e. a feared possible self. Her comment about learning a language to keep one’s mind active - “it’s something you read everywhere, you know? ” - implies that there is also something extrinsic, less internalised, to her foreign language lear‐ ning. It appears that, in line with the prevailing active ageing sentiment in in‐ dustrialized countries (see chapter 2), Christine is a young-old learner who thinks she is ought to commit to lifelong learning, more specifically foreign language learning in order to keep her mind active and to prevent a feared possible self (q.v. Adam, Anne, and Dorothea in chapter 6). Moreover, Dörnyei (2009: 29) admitted that the third component of his model, L2 learning experiences, is “conceptualised at a different level from the two 224 7. Numbers and Words: Merging QUAN and QUAL Strands self-guides and future research will hopefully elaborate on the self aspects of this bottom-up process.” Connected to this, Taylor (2011: 56 f.) further critically remarked that his model is missing an “actual L2 self ” position from which language learners depart towards their ideal L2-selves, and asks “how can we help our students resolve the discrepancy between their present and their future L2 self, if we do not know much about their present L2 self ? ” In what ways can the results of my study further shed light on the L2 experience component of Dörnyei’s model, then? The above discussion has clarified how past and actual L2 selves impact future directed L2 selves or L2 vision. Based on a two-part notion of time, like the one applied in the first round of interview coding (see chapter 6.3.2.1), the young-old temporal self-concept can be thought of consis‐ ting of an actual L2 self that all past and momentary L2-relevant experience have fed into, and a future - target-oriented or guiding - L2 self. The future L2 self can be further differentiated into ideal and ought-to L2 self in accordance with Dörnyei’s self-system. Concluding from this, past and actual L2 selves corre‐ spond with Dörnyei’s L2 learning experience. This study - with both survey and interview - has shed light on how the past and momentary L2 learning experience form the foundation for young-old learners L2-vision or, in reference to Taylor’s question above, the point of departure towards their future L2 selves. The discussion in this chapter underscored once more that approaching the research issue with a mixed methods design was justified in that convergent and divergent results above have provided a more comprehensive picture of young-old language learners. The next chapter will draw conclusions based on the findings of this study and discuss the implications for educational institu‐ tions, instructors, and SLA research. 225 7.2 Discussion in View of Research Questions 8. Conclusion Chapters 5 and 6 have looked at the young-old language learner self and its temporal facets from quantitative and qualitative angles with an eventual mer‐ ging stage in mind. Chapter 7 has brought together the two methodological standpoints to reveal, as well as discuss, points of conand divergence in the data strands. What has crystallized out of the data analyses in those three chap‐ ters is a clearer picture of how young-old language learners - between 60 and 80 years of age and retired - see themselves as language learners in the past, present, and future. Among the recurring themes and trends were keeping one’s mind active (with an underlying “active ageing” rationale), the importance of the social sphere inside the classroom, as well as beyond the classroom in the form of a connection to the L2-community. Furthermore, this study has given a better understanding of the young-old learners’ L2 learning experience, which is comprised of their past and actual L2 selves. It is the aim of this chapter to ‘translate’ the results of this research study into implications for educators and researchers. While the results of this mixed me‐ thods study stem from responses of 195 survey participants and 21 interviewees, readers (including educators and researchers) will need to consider their own, unique classroom setting or research context. As discussed in chapters 2 and 4, what characterizes both young-old language learners and vhs centres is their heterogeneity, the former through their increasing age, the latter due to the need to adapt to local needs and demand, as well as the heterogeneous educational training backgrounds of its language instructors. In other words, this research study does not claim to be representative of and applicable in its entirety to all young-old language learners in Germany. This research study, however, intends to offer grounds for a purposive transferability to other contexts. Moreover, it is hoped that these implications contribute to the field of “foreign language geragogy” (Berndt 2003: 38; Ramírez Gómez 2016) as an evolving sub-division of SLA research. This final chapter will close with further questions that are worth investigating through future research. 8.1 So, What? - Implications for the Foreign Language Classroom 8.1.1 Implications for Instructors Naturally, the implications for young-old learners discussed in this section may also apply to younger or young learners. Nevertheless, these implications stem from the young-old specific characteristics revealed through the empirical re‐ search of this study and therefore - in view of the little existing research on older language learners - need to be emphasised for instructors in adult educa‐ tion contexts. As the analysis of the interviews in chapter 6 has shown, the instrumental motivation of young-old learners has moved away from being a professional one. Two other motives gain in importance instead as young-old learners leave the workforce and continue or start to learn English: defying age-related deficits (in form of ought-to or feared selves) and the social sphere of English classes. In a similar vein, Berndt (2003: 219) writes that educators of older (language) learners have what she calls an “interventionist mission”, which includes both the prevention of age-related deficits and the maintaining of social contacts to prevent age-related isolation. Among other possible consequences of the motivational shift in old age is an increasing intended effort or investment in learning the language - which has been shown to be crucial for a strong L2-vision in the qualitative data analysis. This may be due to the fact that nothing is at stake (e.g. professionally) if young-old learners fail to succeed in their language learning endeavour. Thus, instructors need to consider this shift of motives in relation to learning the language. Both motives should be addressed inand outside the classroom, as the following discussion will exemplify. In terms of the age-defying motives, instructors should firstly raise awareness for this aspect of young-old’s language learning experience and discuss with them what and how this could be implemented. The interviews in chapter 6, however, pointed out that this implementation needs to take account of the young-old’s wish to be treated as grown-up learners and be respected in their self-determination. Keywords such as “homework” or “test” have shown to often unlock negative connotations of being a young student at school again. Thus, language learning activities outside the classroom that enhance the memory training or mind-activating experiences should be negotiated and adapted to individual learners’ preferences and possibilities with regard to their everyday lives. This also means accepting their decision to restrict their language learning to 90 minutes of English class per week. Interviewees who practiced learning English outside the classroom mentioned activities such as reading (books or 227 8.1 So, What? - Implications for the Foreign Language Classroom magazines), listening to English-speaking radio, watching films in English, doing crossword puzzles, helping their (grand-)children with their homework, and sticking lists or cards with new vocabulary to places in their home they would look at often. The fact that not many young-old interviewees were ma‐ king use of these ways of learning or that some were only familiar with more “traditional” ways of memorizing new words shows that there is a need to expand their language learning strategy pool. Familiarizing young-old lear‐ ners, who do not know how to implement activities which appeal to them and do not ‘feel’ like traditional homework or tests, with strategies they can incor‐ porate in their everyday routine should be a goal of English instructors at vhs centres. Once instructors and young-old learners have agreed on activities inand outside the classroom that would further help them train their memory and have implemented this in the foreign language instruction, it is crucial to make lear‐ ners aware of their progress in that regard (additionally to their language lear‐ ning progress). Even though survey respondents and some interviewees noted progress in their language learning, interviewees with a weak L2-vision in par‐ ticular appear to have stopped noticing changes in their memory and language skills. It is this lack of progress - or not noticing any progress - that differen‐ tiates weak L2-vision learners from strong L2-vision learners in this study. Young-old learners with a weak L2-vision experience a conflict of priorities: on the one hand, learning English at vhs centre for enjoyment or with a relaxed approach, and, on the other hand, a need to see a development of their English skills (which quite possibly will require extra effort outside the classroom). Thus, strategies to make learners aware of progress or development of their English skills (as well as their overall cognitive abilities or memory) could help motivate them - this is even more important for young-old learners who have spent several decades at the vhs centre and show signs of a weak L2-vision. Dörnyei (2009: 21) suggests equipping learners with a “‘package’, consisting of an imagery component and a repertoire of appropriate plans, scripts and self-re‐ gulatory strategies” in order to effectively strengthen learners’ L2-vision. This is even more important with young-old learners, who have been exposed to very different, limited or non-existent foreign language instruction at school (in the case of the young-olds in this study this was approximately in the 1940s to 1960s), who, as discovered in the analysis of the survey and interviews, may need to be equipped with more effective learning strategies. One example of achieving this could be to work with portfolios in English classes at vhs centres. One of the advantages of portfolio-work for young-old learners is that it can be adapted to their interests and needs of older learners. Furthermore, young-old 228 8. Conclusion learners get a better sense of their language learning progress as their portfolio grows over semesters and years, in this way also strengthening their L2-vision. The analysis of both data strands has also shown that young-old learners appreciate the social aspect of learning a language. Similar findings in other research studies were presented in chapter 2.2 (Murray 2011; Kade 2009; Ramírez Gómez 2016). For some it is the learning situation itself, i.e. joining a class of like-minded people roughly the same age, for others it is also socializing through using the language abroad. As mentioned in chapter 7, the social aspect has not been explored in detail in the survey. The question about their reasons for lear‐ ning English, however, showed that “learning together with others” is an im‐ portant motive for almost two-thirds of the young-old sample. In contrast to this, the interview data analysis has revealed details about why and how the social sphere of learning a foreign language is so important to young-old lear‐ ners. Wolfram, for instance, mentioned that his social circle became smaller after he retired a few years ago, and that the English group is one way to maintain social contacts in other domains of his life. What does this mean for instructors? Fostering the classroom community needs to be a priority in this adult education context. In class, this entails welcoming and integrating new members with, for example, team-building activities. One interviewee remarked that she changed classes because she did not feel very integrated in her former class and the way conversations were guided or structured in class did not allow every participant - new or old - equal opportunities to speak. It also means reserving time (par‐ ticularly at the beginning of a lesson, when learners come together again after not having seen each other for a week) for chatting about their everyday lives. As one interviewee reported, an initial round of asking “how are you today? ” or “how was your week? ” in combination with a choice of smileys or emotion cards (for example “happy”, “tired”, “excited”, “angry”) to choose from first, gives all participants the opportunity to talk about themselves. It also gives the in‐ structor an impression of the overall mood of individual learners or the whole class (e.g. if several learners had a stressful week or feel tired). Teachers should incorporate these experiences and goals to create a greater awareness of lang‐ uage learner self and vision (or ideal L2 self) in class. Similarly, Berndt (2003: 220) suggests that L2 educators need to place greater emphasis on existent knowledge structures (experiences, wisdom, the actual self) with older learners in contrast to younger learners. A well-balanced conversational structure that alternates between individuals telling something to the rest of the class, and group or partner work, should be the goal for English classes. However, with young-old language learners, whose hearing may be impaired to some extent 229 8.1 So, What? - Implications for the Foreign Language Classroom (see chapter 2 on biological dimension of ageing), work in pairs or groups may make it more difficult to understand what others are saying when there is an increased background noise from other people talking with partners or in groups. Several interviewees have also, for instance, mentioned more relaxed get-to‐ gethers for lunch or dinner at the beginning or end of the semester to cele‐ brate, “field trips” to the theatre to see plays in English, or even trips to England as memorable experiences or highlights during their time at the vhs centre. This means that outside of class, different methods can also foster the classroom community and give young-old learners more opportunities to socialize with other learners beyond the classroom (Benson & Reinders 2011; Legutke 2015) and, in this way, keep learners in class long-term. This also includes re-consi‐ dering the learner-teacher relationship: Finding a balance between a professi‐ onal relationship and accompanying young-old learners on their language lear‐ ning experience rather than leading them top-down as well as the contact or relationship beyond the classroom need to be considered by instructors. In this way, instructors at vhs centres are walking a tight rope between being a teacher and a social worker. Either way, however, and as pointed out by other resear‐ chers (see chapter 2.2) the actual main reason young-old learners come to class should not be left out of sight when organizing get-togethers outside the class‐ room. A connection to learning and using English needs to be made in prepa‐ ration of social events such as going to the theatre, going to a pub after class, or organizing a lunch to celebrate the end of the semester. A characteristic that learners with a strong L2-vision have shown to have in common is the connection to an L2-community where they can use their English skills. Interviewee Tanja, for instance, is motivated whenever she notices con‐ nections between what she learns in class and the “real world”, i.e. the L2-com‐ munity (see chapter 6.4.2.3.1). It seems to give these learners a concrete purpose or goal to work towards to and in this way invest time inand outside the classroom to develop their language skills. How can this be applied to foreign language instruction? In relation to the previously mentioned social events out‐ side class, going to events or places that are conducted or designed in the target language (e.g. a play or a museum tour in English, inviting native speakers etc.) can be one way to increase all young-old learners’ awareness of the relevance or connection to the L2-community in their immediate environment. This is particularly important to those young-old learners who do not have contacts in an L2-community (e.g. relatives) or who are not able to travel anymore (e.g. due to financial or physical limits). This could also include using digital possibilities to build online connections to an L2-community, if young-old learners are open 230 8. Conclusion to this medium of communication and if the learning institution (e.g. the vhs centre) provides the technical infrastructure needed. 8.1.2 Implications for the Institutional Level: The vhs Centres What are the implications for the educational context or institution which is, in the case of this study, the vhs centres in Germany? Something that has been mentioned by interviewees and that should be considered by educational insti‐ tutions is the relatively high fluctuation of teachers in some of the courses. In some cases, teachers were changing every semester or every year, making it hard for young-old learners to re-adjust to new teaching personalities or styles, often resulting in them leaving the course. Moreover, often changing teachers may make it even harder for young-old learners to keep track of their develop‐ ment or get a “coherent sense” of their language learning progress. One of the main reasons for the frequent changes of teachers are the relatively low wages for teachers at vhs centres (see chapter 4). Many foreign language instructors at vhs centres teach there as an additional income, i.e. it is often not a main source of income they depend on. Connected to this, the language instructors’ training backgrounds vary greatly as some have a university degree for teaching languages (often, however, their training has solely been tailored to teaching children and adolescents), some have acquired foreign language skills through their job or longer stays abroad and learn how to teach through short intensive workshops on the weekend, and others again are given a chance to teach as they are native speakers. Only recently, vhs centres introduced the intensive works‐ hops mentioned above which all new language instructors had to complete to provide a more uniform foundation of teaching knowledge and skills for vhs centre teachers (see chapter 4). This is also in line with the European Commis‐ sion’s demand to ensure the “quality of adult learning” (2006: 6 f.). While the intensive workshops are undoubtedly a good start to improving the quality of vhs centre courses and the situations of teachers, a more long-term approach is also needed in addition. As mentioned in chapter 4, vhs instructors often have no place and few opportunities to exchange experiences and learn from each other (e.g. no common teachers room). Additionally, increasing the wages would attract more teachers to stay long-term. This, however, would also mean an increase in course fees and thus possibly make it harder for young-old learners (especially in view of the increase of poverty in old age) on low pensions to afford lifelong learning. Yet, one of the strengths of vhs centres in Germany is that they are - due to their high number of branches in urban and rural areas - able to adapt to the 231 8.1 So, What? - Implications for the Foreign Language Classroom local needs and interests of people living in particular areas. As chapter 2 has outlined, there is an increased variability of individuals as they age; this study’s young-old learners grew older with a tripartite course of life and, connected to this, had certain expectations regarding their retirement, i.e. a new freedom to do something for themselves (in this case, learning or brushing up on a foreign language in order to, for example, to keep their minds active and maintain or extend their social circle). In this way, young-old learners - more than younger learner groups - may need greater variability in course offers (Schmidt-Hertha 2014; Kade 2009). This also applies to foreign language classes that need to meet different demands or interests of older language learners. For instance, the re‐ sults of the study have shown that, compared to their younger counterparts, older language learners may require more opportunities to socialize through the course offers (as discussed in the previous section). When adapting their course offers to local demand, vhs centres should implement and emphasize the soci‐ alizing aspects of their courses and facilities. Similarly, the active ageing reaso‐ ning behind the young-old language learners’ decision to take up language learning at a vhs centre should be foregrounded in the course programme and course design. Some vhs centres in Germany have, for example, started to offer English classes for participants with first signs of dementia. In line with Berndt’s (2003) suggestion for an interventionist mission mentioned above, vhs centres in Germany should make informed decisions regarding how they can attract more young-old learners this way. Considering this, and the results of this study, it is arguable if there should be an overall (e.g. nation-wide) curriculum which is specifically designed for young-old learners, i.e. one that integrates considerations with regard to com‐ petencies that are particularly relevant for older language learners post-retiring. From the previous and the discussion in chapter 4 (research context) it appears that it is an economically wiser decision if vhs centres kept their individual freedom - and, in this way, also their instructors’ freedom - to adapt to the local demand to prevail among competing language course providers which are either in their area or online. 8.2 Beyond this Study: New Questions, Further Research As seen in earlier chapters of this book, this study provides in many ways a starting point for research on the young-old and their language learner self-con‐ cept. It is hoped that my study contributes to a more differentiated view of adult learners in SLA research and that it adds to the further implementation of life‐ 232 8. Conclusion long learning throughout the course of life. Furthermore, in light of the study’s results, employing a mixed methods design helped to find answers to the above research questions and to understand the complexities of young-old language learners’ selves and should thus also be considered for future research on the language learner self. While this study has provided answers to my research questions, it has also uncovered new aspects of older language learners and self-concept research that need further investigation. In view of increasing longevity in industrialized countries and the fact that older people are independent for longer after they retire, it would be interesting to extend the research on older language learners from the young-old to the old-old age group (i.e. learners aged 80 years and older) to see if - along with statistically significant changes in terms of biological age - there are new de‐ velopments in the learner self that need to be considered in language teaching. Similarly, future cohorts of young-old language learners may have different ex‐ periences, needs or expectations regarding their language learning. The young-old participants at the heart of this study are mostly between 60 and 80 years of age and retired, i.e. they were born between 1934 and 1954. As discussed in chapter 2, however, this means that they were also greatly impacted by their time and developed cohort-specific characteristics. One example for the diffe‐ rences between cohorts and generations of young-old is the use of technology. It will be of great interest to explore the current “Gen Y”, who grew up with technology from early stages in their lives and find out to what extent institu‐ tions such as vhs centres and instructors will need to re-think classroom setting (Legutke 2015) and combine the social aspects of language learning with digital possibilities. The social sphere, which has been of central importance to the young-old in this study, may or may not play such a crucial role for future young-old cohorts. This remains something for future SLA research to explore. Connected to the discussion around future cohorts of young-old language learners, however, is also the issue of the future role of lifelong (language) lear‐ ning in one’s life course. As detailed in chapter 2, the current cohorts of young-old learners grew up with a tripartite life course, expecting a period of freedom from familial or work obligations after their working life, i.e. retire‐ ment. In view of current debates around the decreasing pension rates of future generations together with an increasing age for entering retirement, we need to reconsider if the existing lifelong learning initiatives and structures are enough to cater for future young-old learners. Going further, if young-old as in this study, are defined not only by a chronological age-span but also by the sociological age dimension marker of “retirement”, then there will be a defini‐ tional shift of what makes people “young-old” if future generations will start to 233 8.2 Beyond this Study: New Questions, Further Research live a “bipartite” life course in the case of eliminating or transforming the reti‐ rement concept. Extending the time people must work throughout their entire life and keeping them employable in a world with increasing “perforated” wor‐ king lives (i.e. the phenomenon of more frequent job changes) will be an im‐ portant future task for lifelong learning institutions, such as vhs centres in Ger‐ many. With regard to research on the language learner self, this study has provided new perspectives on how older language learners perceive their past, actual, and future selves. Interestingly, their great amount of accumulated experiences with learning and using the language in the past - inside and outside the classroom - has shaped the young-olds’ actual as well as their future L2 self. The latter has been used to group young-old interviewees according to the strength of their outlook on their future language learning development (i.e. L2-vision; see chapter 6). A question that has been raised in the earlier theoretical foundation (chapter 3) but that also recurred in later data analysis stages (for example chapter 6) is the role of socializers in young-old language learners’ lives and the impact or existence of an “ought-to L2 self ” on their self-concepts. Most research on the language learner self has been undertaken with children, adolescents, and young adults at universities - all of whom are still very much influenced in their language study choices by their parents, peers, and teachers. However, as we grow older, what influence do these people - socializers or significant others - have on our plans or motivation to learn a foreign language? On the one hand, the young-old study participants in this have displayed a strong sense of self-de‐ termination and confidence when it comes to the way they want to learn English at the vhs centres. On the other hand, there appear to be remnants of former ought-to L2 selves making young-old learners feel conscious that they should be doing their homework or practicing more outside the classroom. Future re‐ search on older language learners should, therefore, further explore this realm of the language learner self and the implications for foreign language teaching. Additionally, based on the implications from the previous sections, future research could investigate the effectiveness of activities (interventions) that are aimed at fostering L2-vision in older language learners. The results of this study helped to create a picture or describe the status-quo of young-old language le‐ arners’ L2-vision. This study also tried to explain possible reasons and conse‐ quences of a differing L2-vision in young-old learners. The next step for SLA research should therefore be to find ways of working with and improving that L2-vision. Longitudinal case studies that follow older language learners who, for instance, work with portfolios on a long-term basis, can shed light on how their L2-vision is possibly strengthened or how it becomes more defined over 234 8. Conclusion time. Implementing the above suggestions directed at instructors, institutions, and researchers is a first step towards recognizing a differentiated view of adult learners and old age. More importantly, it is a first step to rewarding the incre‐ ased resilience and courage of young-old learners like Jennifer, quoted at the beginning of this thesis, with an enhanced and more fruitful language learning experience. 235 8.2 Beyond this Study: New Questions, Further Research Works Cited Alley, Dawn; Crimmins, Eileen (2010). Epidemiology of Ageing. In Dale Dannefer & Chris Phillipson (Eds.), The SAGE Handbook of Social Gerontology (pp. 75-95). Thousand Oaks: SAGE. Alwin, Duane F. (2010). Social Structure, Cognition, and Ageing. In Dale Dannefer & Chris Phillipson (Eds.), The SAGE Handbook of Social Gerontology (pp. 265-279). Thou‐ sand Oaks: SAGE. Ambos, Ingrid; Dollhausen, Karin; Enders, Kristina; Gnahs, Dieter; Strauch, Anne (2010). Strukturen und Entwicklungen in der Weiterbildung - Ergebnisse der DIE Trendana‐ lyse 2010. DIE Fakten. Ammon, Bridget V.; Maehr, Martin L. (2008). The Resilient Self. Lessons Educators Can Take Away from Risk and Resilience Research. In Dennis M. McInerney, Rhonda G. Craven & Herbert W. Marsh (Eds.), Self-Processes, Learning, and Enabling Human Po‐ tential: Dynamic New Approaches (pp. 289-304): Information Age Publishing. Andrew, Patricia (2012). The Social Construction of Age - Adult Foreign Language Lear‐ ners. Bristol: Multilingual Matters. Aspin, David N.; Chapman, Judith; Evans, Karen; Bagnall, Richard (Eds.). (2012). Second International Handbook of Lifelong Learning - Part One. London: Springer. Bagley, Sylvia S.; Rust, Val D. (2009). Community-Based Folk High Schools in Norway, Sweden, and Denmark. In Rosalind L. Raby & Edward J. Valeau (Eds.), Community College Models. Globalization and Higher Education Reform (pp. 279-298): Springer. Baltes, Paul B.; Smith, Jacqui (2003). New Frontiers in the Future of Aging: From Suc‐ cessful Aging of the Young Old to the Dilemmas of the Fourth Age. Gerontology, 49, 123-135. Bandura, Albert (2001). Social Cognitive Theory: An Agentic Perspective. Annual Review of Psychology, 52(1), 1-26. Bandura, Albert (2012). On the Functional Properties of Perceived Self-Efficacy Revisited. Journal of Management, 38(1), 9-44. Bastian, Hannelore; Meisel, Klaus; Nuissl, Ekkehard; von Rhein, Antje (2004). Kursleitung an Volkshochschulen. Bielefeld: WBV. Benson, Phil; Reinders, Hayo (Eds.). (2011). Beyond the Language Classroom. Basingstoke: Palgrave Macmillan. Bergman, Manfred M. (2008). The Straw Men of the Qualitative-Quantitative Divide and their Influence on Mixed Methods Research. In Manfred M. Bergman (Ed.), Advances in Mixed Methods Research (pp. 1-21): Sage. Berndt, Annette (2003). Sprachenlernen im Alter - Eine empirische Studie zur Fremdspra‐ chengeragogik. München: Iudicum. Bortz, Jürgen; Döring, Nicola (2006). Forschungsmethoden und Evaluation für Human- und Sozialwissenschaftler (4th ed.): Springer. Boyatzis, Richard E.; Akrivou, Kleio (2006). The Ideal Self as the Driver of Intentional Change. Journal of Management Development, 25(7), 624-642. Brandstädter, Jochen; Greve, Werner (1994). The Aging Self: Stabilizing and Protective Processes. Developmental Review, 14, 52-80. Brandstädter, Jochen; Wentura, Dirk; Greve, Werner (1993). Adaptive Resources of the Aging Self: Outlines of an Emergent Perspective. International Journal of Behavioral Development, 16(2), 323-349. Bromley, Dennis B. (1990). Behavioural Gerontology - Central Issues in the Psychology of Ageing. New York: Wiley. Bryman, Alan (2008). Why do Researchers Integrate/ Combine/ Mesh/ Blend/ Mix/ Merge/ Fuse Quantitative and Qualitative Research? In Manfred M. Bergman (Ed.), Advances in Mixed Methods Research (pp. 86-101). London: Sage. Burwitz-Melzer, Eva; Steininger, Ivo (2016). Inhaltsanalyse. In Daniela Caspari; Friederike Klippel; Michael K. Legutke; Karen Schramm (Eds.), Forschungsmethoden in der Fremd‐ sprachendidaktik: Ein Handbuch (pp. 256-268). Tübingen: Narr/ Francke/ Attempto. Byrne, Barbara M. (1996a). Measuring Self-Concept Across the Life Span: Issues and Inst‐ rumentation. Washington, D.C.: APA. Byrne, Barbara M. (1996b). Academic Self-Concept: Its Structure, Measurement, and Re‐ lation to Academic Achievement. In Bruce A. Bracken (Ed.), Handbook of Self-Concept. Developmental, Social, and Clinical Considerations. (pp. 287-316). New York: Wiley. Byrne, Barbara M.; Shavelson, Richard J. (1996). On the Structure of Social Self-Concept for Pre-, Early, and Late Adolescents: A Test of the Shavelson, Hubner, and Stanton (1976) Model. Journal of Personality and Social Psychology, 70(3), 599-613. 237 Works Cited Candlin, Christopher N.; Riazi, Mehdi (2014). Mixed-Methods Research in Language Tea‐ ching and Learning: Opportunities, Issues and Challenges. Language Teaching, 47(02), 135-173. Canfield, Jack; Wells, Harold C. (1976). A 100 Ways to Enhance Self-concept in the Class‐ room. New Jersey: Prentice Hall. Campbell, Donald T.; Fiske, Donald W. (1959). Convergent and Discriminant Validation by the Multitrait-Multimethod Matrix. Psychological Bulletin, 56(2), 81-105. Carstensen, Laura L. (2006). The Influence of a Sense of Time on Human Development. Science, 312(5782), 1913-1915. Cavanaugh, John C.; Blanchard-Fields, Fredda (2002). Adult Development and Aging (4th ed.). Belmont: Wadsworth. Coetzee-Van Rooy, Susan (2006). Integrativeness: Untenable for World Englishes Lear‐ ners? World Englishes, 25(3/ 4), 437-450. Cohen, Louis; Manion, Lawrence; Morrison, Keith (2011). Research Methods in Education (7th ed.). London: Routledge. Cooley, Charles H. (1922). Human Nature and the Social Order. New York: Charles Scrib‐ ner's Sons. Costard, Astrid (2006). Der Übergang in den Ruhestand als Bezugspnkt für Bildungsange‐ bote. (Diploma Thesis), Universität zu Köln, Bonn. Retrieved from www.die-bonn.de/ doks/ costard0601.pdf Creswell, John W. (2009). Mixed Methods Procedures. In John W. Creswell (Ed.), Research Design (3rd ed., pp. 203-225). London: Sage. Creswell, John W.; Ivankova, Nataliya V. (2009). Mixed Methods. In Juanita Heigham; Robert A. Croker (Eds.), Qualitative Research in Applied Linguistics: A Practical Intro‐ duction (pp. 135-161). Basingstoke: Palgrave Macmillan. Creswell, John W.; Plano Clark, Vicki L. (2011 [2007]). Designing and Conducting Mixed Methods Research. Thousand Oaks: Sage. Creswell, John W.; Plano Clark, Vicki L.; Garrett, Amanda (2008). Methodological Issues in Conducting Mixed Methods Research Designs. In Manfred M. Bergman (Ed.), Ad‐ vances in Mixed Methods Research (pp. 66-83). Thousand Oaks: Sage. Data Analysis Australia (DAA) (2016). Response Rates. Retrieved from www.daa.com.au/ analytical-ideas/ response-rates/ 238 Works Cited Daatland, Svein O. (2005). Quality of Life and Ageing. In Malcolm L. Johnson (Ed.), Age and Ageing (pp. 371-377). Cambridge: CUP. Dahlberg, Britt; Wittink, Marsha N.; Gallo, Joseph J. (2010). Funding and Publishing In‐ tegrated Studies. Writing Effective Mixed Methods Manuscripts and Grand Proposals. In Abbas Tashakkori; Charles Teddlie (Eds.), Sage Handbook of Mixed Methods in Social & Behavioral Research (2nd ed., pp. 775-802). Thousand Oaks: Sage. Dannefer, Dale; Settersten, Richard A. J. (2010). The Study of the Life Course: Implications for Social Gerontology. In Dale Dannefer; Chris Phillipson (Eds.), The Sage Handbook of Social Gerontology (pp. 5-19). Thousand Oaks: Sage. Deci, Edward L.; Ryan, Richard M. (1985). The General Causality Orientations Scale: Self-Determination in Personality. Journal of Research in Personality, 19, 109-134. Doff, Sabine (2005). Zentrale Faktoren der Lernerindividualität Erwachsener: Empirische Analyse und Plädoyer für einen adressatenbezogenen Fremdsprachenunterricht. In Rüdiger Ahrens; Ursula Weier (Eds.), Englisch in der Erwachsenenbildung des 21. Jahr‐ hunderts. (pp. 205-228). Heidelberg: Winter. Dorfman, Lorraine T.; Kolarik, Douglas C. (2005). Leisure and the Retired Professor: Oc‐ cupation Matters. Educational Gerontology, 31, 343-361. Dörnyei, Zoltán (2007). Research Methods in Applied Linguistics - Quantitative, Qualita‐ tive, and Mixed Methodologies. Oxford: OUP. Dörnyei, Zoltán; Taguchi, Tatsuya (2010). Questionnaires in Second Language Research. Construction, Administration, and Processing (2nd ed.). New York: Routledge. Dörnyei, Zoltán (2005). The Psychology of the Language Learner. Individual Differences in Second Language Acquisition. New Jersey: Lawrence Erlbaum Associates. Dörnyei, Zoltán (2009). The L2 Motivational Self System. In Zoltán Dörnyei; Ema Ushioda (Eds.), Motivation, Language Identity and the L2 Self (pp. 9-42). Bristol: Multilingual Matters. Dörnyei, Zoltán; Ushioda, Ema (2011). Teaching and Researching Motivation (2nd ed.). London: Pearson. Dresing, Thorsten; Pehl, Thorsten, Schmieder, Christian (2015). Manual (on) Transcrip‐ tion. Transcription Conventions, Software Guides and Practical Hints for Qualitative Re‐ searchers Retrieved from www.audiotranskription.de/ english/ transcription-practicalguide.htm 239 Works Cited Elder, Glen H. J. (1974). Children of the Great Depression: Social Change in Life Experi‐ ence. Chicago: CUP. Elder, Glen H. J.; Kirkpatrick Johnson, Monica; Crosnoe, Robert (2003). The Emergence and Development of Life Course Theory. In Jeylan T. Mortimer; Michael J. Shanahen (Eds.), Handbook of the Life Course (pp. 3-19). New York: Kluwer/ Plenum. Erikson, Martin G. (2007). The Meaning of the Future: Toward a More Specific Definition of Possible Selves. Rev. Gen. Psychol., 11(4), 348-358. Eschmann, Detlef (2005). Englischlernende an Volkshochschulen. In Rüdiger Ahrens & Ursula Weier (Eds.), Englisch in der Erwachsenenbildung des 21. Jahrhunderts (pp. 327-338). Heidelberg: Winter. European Commission (2006). Adult Learning: It is never too late to learn. Retrieved from https: / / eur-lex.europa.eu/ legal-content/ EN/ TXT/ PDF/ ? uri=CELEX: 52006DC0614&from=EN European Commission (2009). Key competences for a changing world. Draft 2010 joint progress report of the Council and the Commission on the implementation of the "Education and Training 2010 work programme" Retrieved from https: / / eur-lex.europa.eu/ legal-content/ EN/ TXT/ PDF/ ? uri=CELEX: 52009DC0640&from=DE European Commission (14th January 2016). Adult Learning. Retrieved from http: / / ec.europa.eu/ education/ policy/ adult-learning/ adult_en.htm Field, Andy (2009). Discovering Statistics Using SPSS (3rd ed.). Los Angeles: Sage. Flick, Uwe (2011). Triangulation - Eine Einführung (3rd ed.). Wiesbaden: Verlag für So‐ zialwissenschaften. Fodor, Jerry (1998). The Trouble with Psychological Darwinism. London Review of Books, 20(2), 11-13. Friebe, Jens (2007). Weiterbildung in einer alternden Gesellschaft. Bestandsaufnahme der demografischen Entwicklungen, des Lernens im höheren Lebensalter und der Per‐ spektiven für die Weiterbildung. Retrieved from www.die-bonn.de/ doks/ friebe0602.pdf Friebertshäuser, Barbara (1997). Interviewtechniken - ein Überblick. In Friebertshäuser, Barbara et al. (Eds.), Handbuch Qualitativer Forschungsmethoden in der Erziehungs‐ wissenschaft (pp. 371-395). Weinheim: Juventa. Gay, L. R.; Mills, Geoffrey E.; Airasian, Peter (2012). Educational Research - Competencies for Analysis and Applications (10th ed.). Boston: Pearson. 240 Works Cited Gilleard, Chris; Higgs, Paul (2002). The Third Age: Class, Cohort or Generation? Ageing and Society, 22(3), 369-382. Gillham, Bill (2000). Developing a Questionnaire. London: Continuum. Gläser-Zikuda, Michaela (2011). Qualitative Auswertungsverfahren. In Heinz Reinders et al. (Eds.), Empirische Bildungsforschung (pp. 109-119). Wiesbaden: Verlag für Sozi‐ alwissenschaften. Greenberg, Joseph A. (1991). Exporting the Community College Concept: Worldwide Variations. New Directions for Community Colleges, 75, 69-77. Greene, Jennifer C.; Caracelli, Valerie J.; Graham, Wendy F. (1989). Toward a Conceptual Framework for Mixed-Method Evaluation Designs. Educational Evaluation and Policy Analysis, 11(3), 255-274. Gronemeyer, Reimar; Buff, Wolfgang (1992). Bildungskonzepte oder Orientierung der Älteren? In Jürgen Reggentin-Dettbarn; Heike Reggentin (Eds.), Neue Wege in der Bil‐ dung Älterer - Band. 2 Praktische Modelle und Projekte (pp. 21-29). Freiburg: Lambertus. Grum, Urška; Legutke, Michael K. (2016). Sampling. In Daniela Caspari; Friederike Klippel; Michael K. Legutke; Karen Schramm (Eds.), Forschungsmethoden in der Fremd‐ sprachendidaktik - Ein Handbuch (pp. 78-89). Tübingen: Narr/ Francke/ Attempto. Hammersley, Martyn (2008). Troubles with Triangulation. In Manfred M. Bergman (Ed.), Advances in Mixed Methods Research (pp. 22-37). London: Sage. Hammond, Cathie (2005). The Wider Benefits of Adult Learning: An Illustration of the Advantages of Multi-method Research. International Journal of Social Research Me‐ thodology, 8(3), 239-255. Hardy, Melissa (2011). Rethinking Retirement. In Richard A. J. Settersten; J. L. Angel (Eds.), Handbook of Sociology of Aging (pp. 213-227). New York: Springer. Harter, Susan (1996). Historical Roots of Contemporary Issues involving Self-Concept. In Bruce A. Bracken (Ed.), Handbook of Self-Concept (pp. 1-37). New York: Wiley. Harter, Susan; Bouchey, Heather A. (2005). Reflected Appraisals, Academic Self-Percep‐ tions, and Math/ Science Performance During Early Adolescence. Journal of Educati‐ onal Psychology, 97(4), 673-686. Harter, Susane; Whitesell, N. R. (2003). Beyond the Debate: Why Some Adolescents Re‐ port Stable Self-Worth Over Time and Situation, Whereas Others Report Changes in Self-Worth. Journal of Personality, 71(6), 1027-1058. Hattie, John (1992). Self-Concept. New Jersey: Erlbaum. 241 Works Cited Hattie, John; Song, In-Sub (1984). Home Environment, Self-Concept, and Academic Achievement: A Causal Modeling Approach. Journal of Educational psychology, 76(6), 1269-1281. Haustein, Thomas; Mischke, Johanna (2011). Im Blickpunkt. Ältere Menschen in Deutsch‐ land und der EU. Herschensohn, Julia (2007). Language Development and Age. Cambridge: CUP. Higgins, Edward T. (1997). Beyond Pleasure and Pain. American Psychologist, 52(12), 1280-1300. Higgins, Edward T. (1987). Self-Discrepancy: A Theory Relating Self and Affect. Psycho‐ logical Review, 94(3), 319-340. Huntemann, Hella (2014). Volkshochschul-Statistik 2013 Zahlen in Kürze. Retrieved from: www.die-bonn.de/ doks/ 2014-volkshochschule-statistik-02.pdf Huntemann, Hella; Reichart, Elisabeth (2014). Volkshochschul-Statistik: 52. Folge, Arbeits‐ jahr 2013. Retrieved from: www.die-bonn.de/ doks/ 2014-volkshochschule-statistik-01.pdf James, William (1890). The Principles of Psychology: Henry Hold and Company. Jarvis, Peter (2009a). The Routledge International Handbook of Lifelong Learning. (P. Jarvis Ed.). London: Routledge. Jarvis, Peter (2009b). The European Union and Lifelong Learning Policy The Routledge International Handbook of Lifelong Learning (pp. 271-280). London: Routledge. Johnson, Malcolm L. (2005). The Social Construction of Old Age as a Problem. In M. L. Johnson (Ed.), The Cambridge Handbook of Age and Ageing (pp. 563-571). Cambridge: CUP. Jopling, David A. (1997). A "Self or Selves"? In Ulric Neisser (Ed.), The Conceptual Self in Context: Culture, Experience, Self-understanding (pp. 249-267). New York: CUP. Jopling, David A. (2002). Self-Knowledge and the Self. New York: Routledge. Jovic, Emily; McMullin, Juli (2011). Learning and Aging. In Richard A. J. Settersten; J. L. Angel (Eds.), Handbook of Sociology and Aging: Springer. Kade, Sylvia (2007). Altern und Bildung. Eine Einführung. Bielefeld: WBV. Kade, Sylvia (2009). Altern und Bildung - Eine Einführung (2nd ed.). Bielefeld: WBV. Karl, Fred (2009). Einführung in die Generationen- und Altenarbeit. Opladen: Budrich. 242 Works Cited Käpplinger, Bernd (2014). Adult Education Research in Poland and Germany on the Eu‐ ropean Level: Two Sleeping Beauties? Rocznik Andragogiczny, 21, 393-409. Klein, Stanley B.; Gangi, Cynthia E. (2010). The Multiplicity of Self: Neuropsychological Evidence and Its Implications for the Self as a Construct in Psychological Research. Annals of the New York Academy of Sciences, 1191(The Year in Cognitive Neuroscience), 1-15. Klippel, Johanna (2011). Die Herausforderung von leitfadengestützen Interviews als For‐ schungsinstrument in der Fremdsprachenforschung. Zeitschrift für Interkulturellen Fremdsprachenunterricht., 16(2), 81-92. Knorr, Petra; Schramm, Karen (2016). Triangulation. In Daniela Caspari; Friederike Klippel; Michael K. Legutke; Karen Schramm (Eds.), Forschungsmethoden in der Fremd‐ sprachendidaktik - Ein Handbuch (pp. 90-97). Tübingen: Narr/ Francke/ Attempto. Kolland, Franz; Klingenberg, Heinrich (2011). Lebenslanges Lernen im späteren Lebens‐ alter. Grundlagen und Begriffserklärungen. In Waxenegger, Andrea (Ed.), Lernen und Bildung im späteren Lebensalter. Leitlinien und Prioritäten 2020 (pp. 17-32). Graz: Tha‐ lerhof. Komp, Kathrin; Aartsen, Marja (2013). Introduction: Older People under the Magnifying Glass. In Kathrin Komp; Marja Aartsen (Eds.), Old Age in Europe: A Textbook of Ge‐ rontology (pp. 1-13). Dordrecht: Springer. Kormos, Judit; Csizér, Kata (2008). Age-Related Differences in the Motivation of Learning EFL: Attitudes, Selves, and Motivated Learning Behavior. Language Learning, 58(2), 327-355. Kotter-Grühn, Dana; Hess, Thomas M. (2012). The Impact of Age Stereotypes on Self-per‐ ceptions of Aging Across the Adult Lifespan. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 67(5), 563-571. Koutoukidis, Gabby; Stainton, Kate; Hughson, Jodie (2013). Tabbner's Nursing Care: Theory and Practice (6th ed.): Elsevier Australia. Kowal, Sabine; O'Connell, Daniel C. (2014). Transcription as a Crucial Step of Data Ana‐ lysis. In Uwe Flick (Ed.), The SAGE Handbook of Qualitative Data Analysis (pp. 64-78). London: SAGE. Kruse, Andreas (2008). Alter und Altern - Konzeptionelle Überlegungen und empirische Befunde der Gerontologie. In Andreas Kruse (Ed.), Weiterbildung in der zweiten Le‐ benshälfte. Bielefeld: Bertelsmann. 243 Works Cited Kuckartz, Udo (2014). Qualitative Inhaltsanalyse. Methoden, Praxis, Computerunterstüt‐ zung (2nd ed.). Basel: Beltz Juventa. Kulich, Jindra (1964). The Danish Folk High School: Can It Be Transplanted? The Success and Failure of the Danish Folk High School at Home and Abroad. International Review of Education, 10(4), 417-430. L'Ecuyer, René (1975). Self-Concept Investigation: Demystification Process. Journal of Phenomenological Psychology, 6(1), 17-30. Lang, Frieder R.; Martin, Mike; Pinquart, Martin (2012). Entwicklungspsychologie Erwach‐ sener. Göttingen: Hogrefe. Laslett, Peter (1987). The Emergence of the Third Age. Ageing and Society, 7(2), 133-160. Laslett, Peter (1994). The Third Age, The Fourth Age and The Future. Ageing and Society, 14(3), 436-447. Lazaraton, Anne (2005). Quantitative Research Methods. In Eli Hinkel (Ed.), Handbook of Research in Second Language Teaching and Learning (pp. 209-224). New York: Taylor and Francis. Leary, Mark R.; Tangney, June Price (2012). Handbook of Self and Identity (2nd ed.). New York: Guilford Press. Legutke, Michael K. (2015). Vernetzte Lernorte. In Eva Burwitz-Melzer; Frank G. Königs; Claudia Riemer (Eds.), Lernen an allen Orten? - Die Rolle der Lernorte beim Lehren und Lernen von Fremdsprachen (pp. 127-135). Tübingen: Narr. Legutke, Michael K. (2016). Zusammenfassung und Diskussion der Erträge. In Daniela Caspari; Friederike Klippel; Michael K. Legutke; Karen Schramm (Eds.), Forschungs‐ methoden in der Fremdsprachendidaktik - Ein Handbuch (pp. 386-392). Tübingen: Narr/ Francke/ Attempto. Legutke, Michael K.; Schramm, Karen (2016). Forschungsethik. In Daniela Caspari; Frie‐ derike Klippel; Michael K. Legutke; Karen Schramm (Eds.), Forschungsmethoden in der Fremdsprachendidaktik - Ein Handbuch (pp. 108-117). Tübingen: Narr/ Francke/ Attempto. Lehr, Ursula (2006). Demographischer Wandel. In Wolf Oswald (Ed.), Gerontologie - Me‐ dizinische, psychologische und sozialwissenschaftliche Grundbegriffe (pp. 159-164). Stuttgart: Kohlhammer. Lehr, Ursula (2007). Psychologie des Alterns (2nd ed.). Wiebelsheim: Quelle & Meyer. 244 Works Cited Leung, Kim C.; Marsh, Herbert W.; Craven, Rhonda G.; Yeung, Alexander S.; Abduljabbar, Adel S.(2013). Domain Specificity Between Peer Support and Self-Concept. Journal of Early Adolescence, 33(2), 227-244. doi: 10.1177/ 0272431611436130 Macintyre, Peter D.; Mackinnon, Sean P.; Clément, Richard (2009). The Baby, The Bath‐ water, and the Future of Language Learning Motivation Research. In Zoltán Dörnyei; Ema Ushioda (Eds.), Motivation, Language Identity and the L2 Self (pp. 43-65). Bristol: Multilingual Matters. Markus, Hazel; Nurius, Paula (1986). Possible Selves. American Psychologist, 41(9), 954-969. Markus, Hazel; Wurf, Elissa (1987). The Dynamic Self-concept: A Social Psychological Perspective. Annual Review of Psychology, 38(1), 299-337. Marsh, Herbert W. (1986). Verbal and Math Self-concepts: An Internal/ External Frame of Reference Model. American Educational Research Journal, 23(1), 129-149. Marsh, Herbert W. (1990). The Structure of Academic Self-concept: The Marsh/ Shavelson Model. Journal of Educational Psychology, 82(4), 623. Marsh, Herbert W.; Ayotte, Violaine (2003). Do Multiple Dimensions of Self-Concept Become More Differentiated with Age? The Differential Distinctiveness Hypothesis. Journal of Educational psychology, 95(4), 687-706. Marsh, Herbert W.; Byrne, Barbara M.; Shavelson, Richard J. (1988). A Multifaceted Aca‐ demic Self-concept: Its Hierarchical Structure and Its Relation to Academic Achieve‐ ment. Journal of Educational Psychology, 80(3), 366-380. Marsh, Herbert W.; Hattie, John (1996a). Theoretical Perspectives on the Structure of Self-Concept. In Bruce A. Bracken (Ed.), Handbook of Self-Concept. Developmental, Social, and Clinical Considerations (pp. 38-90). New York: Wiley. Marsh, Herbert W.; Hattie, John (1996b). Future Directions in Self-Concept Research. In Bruce A. Bracken (Ed.), Handbook of Self-Concept. Developmental, Social, and Clinical Considerations (pp. 421-462). New York: Wiley. Marsh, Herbert W.; Shavelson, Richard J. (1985). Self-Concept: Its Mulitfaceted, Hierar‐ chical Structure. Educational Psychologist, 20, 107-125. Marshall, Victor W.; Taylor, Philip (2005). Restructuring the Lifecourse: Work and Reti‐ rement. In Malcolm L. Johnson (Ed.), The Cambridge Handbook of Age and Ageing (pp. 572-582). Cambridge: CUP. 245 Works Cited Masogret, Anne-Marie; Gardner, Robert C. (2003). Attitudes, Motivation, and Second Language Learning: A Meta-Analysis of Studies Conducted by Gardner and Asso‐ ciates. Language Learning, 53(1), 123-163. Maxwell, Joseph A. (2005). Qualitative Research Design: An Interactive Approach (2nd ed.). Thousand Oaks: Sage. Mayring, Philipp (2010). Qualitative Inhaltsanalyse. Grundlagen und Techniken (11th ed.). Weinheim: Beltz. Mayring, Philipp (2014). Qualitative Content Analysis. Theoretical Foundation, Basic Pro‐ cedures and Software Solution. Weinheim: Beltz. McCaslin, Mark L. (2008). Pragmatism. In Lisa M. Given (Ed.), The Sage Encyclopedia of Qualitative Research Methods. Thousand Oaks: Sage. McEwon, Maya S.; Noels, Kimberly A.; Chaffee, Kathryn E. (2014). At the Interface of the Socio-Educational Model, Self-Determination Theory and the L2 Motivational Self System Models. In Kata Csizér; Michael Magid (Eds.), The Impact of Self-Concept on Language Learning (pp. 19-50). Bristol: Multilingual Matters. Mead, George H. (1922). A Behavioristic Account of the Significant Symbol. Indianapolis: Bobbs-Merrill. Mercer, Sarah (2011a). Towards an Understanding of Language Learner Self-Concept. Hei‐ delberg: Springer. Mercer, Sarah (2011b). Language Learner Self-concept: Complexity, Continuity and Change. System, 39(3), 335-346. Mercer, Sarah (2014). The Self from a Complexity Perspective. In Sarah Mercer; Marion Williams (Eds.), Multiple Perspectives on the Self in SLA (pp. 160-176). Bristol: Multi‐ lingual Matters. Mercer, Sarah; Williams, Marion (2014). Concluding Remarks. In Sarah Mercer; Marion Williams (Eds.), Multiple Perspectives on the Self in SLA (pp. 177-185). Bristol: Multi‐ lingual Matters. Mertens, Donna M. (2010). Publishing Mixed Methods Research. Journal of Mixed Methods Research, 5(1), 3-6. Merton, Robert K. (1968). The Matthew Effect in Science. Science, 159(3810), 56-63. Midwinter, Eric (2005). How Many People are There in the Third Age? Ageing and Society, 25(1), 9-18. 246 Works Cited Mills, Nicole (2014). Self-Efficacy in Second Language Acquisition. In Sarah Mercer; Ma‐ rion Williams (Eds.), Multiple Perspectives on the Self in SLA (pp. 6-22). Bristol: Multi‐ lingual Matters. Mlinac, Michelle E.; Sheeran, Tom H.; Blissmer, Bryan; Lees, Faith; Martins, Diane (2011). Psychological Resilience. In Barbara Resnick; Lisa P. Gwyther; Karen A. Roberto (Eds.), Resilience in Aging: Concepts, Research, and Outcomes (pp. 67-87). New York: Springer. Moody, Harry R. (2010). The New Ageing Enterprise. In Dale Dannefer; Chris Phillipson (Eds.), The Sage Handbook of Social Gerontology (pp. 483-494). Thousand Oaks: Sage. Morse, Janice M. (2010). Procedures and Practice of Mixed Method Design: Maintaining Control, Rigor, and Complexity. In Abbas Tashakkori; Charles Teddlie (Eds.), Sage Handbook of Mixed Methods in Social & Behavioral Research (2nd ed., pp. 339-352). Thousand Oaks: Sage. Morse, Janice M.; Maddox, Lory J. (2013). Analytic Integration in Qualitatively Driven (QUAL) Mixed and Multiple Methods Designs. In Uwe Flick (Ed.), The SAGE Handbook of Qualitative Data Analysis (pp. 524-539). London: Sage. Mummendey, Hans D. (2006). Psychologie des 'Selbst' - Theorien, Methoden und Ergebnisse der Selbstkonzeptforschung. Göttingen: Hogrefe. Mummendey, Hans D.; Grau, Ina (2008). Die Fragebogen-Methode (5th ed.). Göttingen: Hogrefe. Murray, Denise E. (2005). ESL in Adult Education. In Eli Hinkel (Ed.), Handbook of Re‐ search in Second Language Teaching and Learning (pp. 65-84). New York: Taylor and Francis. Murray, Garold (2009). Narrative Inquiry. In Juanita Heigham; Robert A. Croker (Eds.), Qualitative Research in Applied Linguistics: A Practical Introduction (pp. 45-89). Ba‐ singstoke: Palgrave Macmillan. Murray, Garold (2011). Older Language Learners, Social Learning Spaces and Community. In Phil Benson; Hayo Reinders (Eds.), Beyond the Language Classroom (pp. 132-145). London: Palgrave Macmillan. Nastasi, Bonnie K.; Hitchcock, John H.; Brown, Lisa M. (2010). An Inclusive Framework for Conceptualizing Mixed Methods Design Typologies. Moving Toward Fully Integ‐ rated Synergistic Research Models. In Abbas Tashakkori; Charles Teddlie (Eds.), Sage Handbook of Mixed Methods in Social Behavioral Research (pp. 305-338). Thousand Oaks: Sage. 247 Works Cited Neigert, Miriam (2011). Teaching the Young-Old - A Task-Based Approach. (Diploma Thesis), Justus-Liebig Universität, Gießen. Neisser, Ulric (1988). Five Kinds of Self-knowledge. Philosophical Psychology, 1(1), 35-59. Neisser, Ulric (1993). The Self Perceived. In Ulric Neisser (Ed.), The Perceived Self: Ecolo‐ gical and Interpersonal Sources of Self Knowledge. (pp. 3-21). Cambridge: Cambridge University Press. Neisser, Ulric (1997). Concepts and Self-Concepts. In Ulric Neisser; David A. Jopling (Eds.), The Conceptual Self in Context: Culture, Experience, Self-Understanding. (pp. 3-12). Cambridge: Cambridge University Press. Neugarten, Bernice L. (1974). Age Groups in American Society and the Rise of the Young-Old. The Annals of the American Academy of Political and Social Science, 415 (Political Consequences of Aging), 187-198. Nimrod, Galit; Ben-Shem, Idit (2015). Successful Aging as a Lifelong Process. Educational Gerontology, 41(11), 814-824. Nolda, Sigrid (2008). Einführung in die Theorie der Erwachsenenbildung. Darmstadt: WBG. Northoff, Georg (2014). Brain and Self: A Neurophilosophical Account. In Sarah Mercer; Marion Williams (Eds.), Multiple Perspectives on the Self in SLA (pp. 142-159). Bristol: Multilingual Matters. Nugrahenny, Zacharias T. (2012). Qualitative Research Methods for Second Language Edu‐ cation: A Coursebook. Newcastle: Cambridge Scholars Publishing. Nunan, David (1992). Research Methods in Language Learning. Cambridge: CUP. Nunan, David; Bailey, Kathleen M. (2009). Exploring Second Language Classroom Re‐ search - A Comprehensive Guide. Boston: Heinle. O'Leary, Zina (2014). The Essential Guide to Doing your Research Project (2nd ed.). Thou‐ sand Oaks: Sage. O'Rand, Angela M.; Isaacs, Katelin; Roth, Leslie (2010). Age and Inequality in Global Context. In Dale Dannefer; Chris Phillipson (Eds.), The SAGE Handbook of Social Ge‐ rontology (pp. 127-136). Thousand Oaks: SAGE. Orth, Ulrich; Robins, Richard W. (2014). The Development of Self-Esteem. Current Di‐ rections in Psychological Science, 23(5), 381-387. Oyserman, Daphna (2008). Possible Selves. Identity-Based Motivation and School Suc‐ cess. In Dennis M. McInerney; Rhonda G. Craven; Herbert W. Marsh (Eds.), Self-Pro‐ 248 Works Cited cesses, Learning, and Enabling Human Potential: Dynamic New Approaches (pp. 269-288): Information Age Publishing. Oyserman, Daphna; Elmore, Kristen; Smith, George (2011). Self, Self-Concept and Iden‐ tity. In Mark R. Leary; June P. Tangney (Eds.), Handbook of Self and Identity (2 ed., pp. 69-104). New York: Guilford. Oyserman, Daphna; Markus, Hazel R. (1998). Self as Social Representation. In Uwe Flick (Ed.), The Psychology of the Social (pp. 107-126). New York: CUP. Pajares, Frank; Schunk, Dale H. (2002). Self and Self-Belief in Psychology and Education: A Historical Perspective. In Joshua Aronson (Ed.), Improving Academic Achievement. Impact of Psychological Factors on Education (pp. 3-21). San Diego: Academic Press. Pfirrmann, Monika. EPQ - Erwachsenpädagogische Qualifizierung. Retrieved 5. Sep‐ tember 2015 from www.vhs-in-hessen.de/ Artikel/ cmx4ee8b9bd1588e.html Phillips, Judith; Ajrouch, Kristine; Hillcoat-Nallétamby, Sarah (2010). Key Concepts in Social Gerontology. Thousand Oaks: Sage. Plano Clark, Vicki L.; Badiee, Manijeh (2010). Research Questions in Mixed Methods Research. In Abbas Tashakkori; Charles Teddlie (Eds.), Sage Handbook of Mixed Me‐ thods in Social & Behavioral Research (2nd ed., pp. 275-304). Thousand Oaks: Sage. Pongratz, Ludwig A. (2003). Zeitgeistsurfer - Beiträge zur Kritik der Erwachsenenbildung. Basel: Beltz. Raasch, Albert (2005). "Gut" schließt "besser" nicht aus. Gedanken zur Innovation des Fremdsprachenunterrichts in der Erwachsenenbildung. In Rüdiger Ahrens; Ursula Weier (Eds.), Englisch in der Erwachsenenbildung des 21. Jahrhunderts. (pp. 23-48). Heidelberg: Winter. Raby, Rosalind L. (2009). Defining the Community College Model. In Rosalind L. Raby; Edward J. Valeau (Eds.), Community College Models. Globalization and Higher Educa‐ tion Reform. (pp. 3-19): Springer. Ramírez Gómez, Danya (2016). Critical Geragogy and Foreign Language Learning: An Exploratory Application. Educational Gerontology, 42(2), 136-143. Reinders, Hayo (2011a). Fragebogen. In Hayo Reinders (Ed.), Empirische Bildungsfor‐ schung (pp. 53-65). Wiesbaden: Verlag für Sozialwissenschaften. Reinders, Hayo (2011b). Interview. In Hayo Reinders (Ed.), Empirische Bildungsforschung (pp. 85-97). Wiesbaden: Verlag für Sozialwissenschaften. 249 Works Cited Reutter, Gerhard (2004). Diskontinuierliche Erwerbsbiographien - Realität oder Konst‐ rukt? DIE Fakten. Richards, Keith (2009). Interviews. In Juanita Heigham & Robert A. Croker (Eds.), Qua‐ litative Research in Applied Linguistics: A Practical Introduction (pp. 182-199). Basings‐ toke: Palgrave Macmillan. Rossman, Gretchen B.; Wilson, Bruce L. (1994). Numbers and Words Revisited: Being "Shamelessly Eclectic". Quality and Quantity, 28, 315-327. Roulston, Kathryn (2014). Analysing Interviews. In U. Flick (Ed.), The SAGE Handbook of Qualitative Data Analysis (pp. 297-312). London: SAGE. Rowe, John W.; Kahn, Robert L. (1997). Successful Aging. The Gerontologist, 37(4), 433-440. Rowland, Donald T. (2012). The Third Age. In Donald T. Rowland (Ed.), Population Aging: The Transformation of Societies (pp. 167-181): Springer. Rubio, Fernando D. (2014). Self-Esteem and Self-Concept in Foreign Language Learning. In Sarah Mercer; Marion Williams (Eds.), Multiple Perspectives on the Self in SLA (pp. 41-58). Bristol: Multilingual Matters. Russell, Helen (2011). Time and Meaning in Later-life Learning. Australian Journal of Adult Learning, 51(3), 547-565. Ryan, Stephen; Irie, Kay (2014). Imagined and Possible Selves: Stories We Tell Ourselves About Ourselves. In Sarah Mercer; Marion Williams (Eds.), Multiple Perspectives on the Self in SLA (pp. 109-126). Bristol: Multilingual Matters. Ryff, Carol D (1991). Possible Selves in Adulthood and Old Age: A Tale of Shifting Ho‐ rizons. Psychology and Aging, 6(2), 286-295. doi: 10.1037/ 0882-7974.6.2.286 Ryle, Gilbert (1949). The Concept of Mind. Chicago: University of Chicago Press. Sampasivam, Sinthujaa & Clément, Richard (2014). The Dynamics of Second Language Confidence: Perspective and Interaction. In Sarah Mercer; Marion Williams (Eds.), Multiple Perspectives on the Self in SLA (pp. 23-40). Bristol: Multilingual Matters. Schaie, K. Warner (2015). Theoretical Perspectives for the Psychology of Aging in a Life‐ span Context. In K. Warner Schaie; Sherry L. Willis (Eds.), Handbook of the Psychology of Aging (8th ed., pp. 3-13). San Diego: Elsevier. Schart, Michaela (2003). Projektunterricht - subjektiv betrachtet. Eine qualitative Studie mit Lehrenden für Deutsch als Fremdsprache. Hohengehren: Schneider Verlag. 250 Works Cited Schmidt, Elke-Heidrun (2005). "Generation 50plus" - kommerzielle Erfindung oder neue Zielgruppe für die Erwachsenenbildung? Retrieved from www.die-bonn.de/ esprid/ dokumente/ doc-2005/ schmidt05_01.pdf Schmidt-Hertha, Bernhard (2014). Kompetenzerwerb und Lernen im Alter. Bielefeld: WBV / DIE. Schnurr, Simone; Theisen, Catharina (2009). Soziale Netzwerke und Familie. In Rudolf Tippelt, Bernhard Schmidt; Simone Schnurr; Simone Sinner; Catharina Theisen (Eds.), Bildung Älterer - Chancen im demografischen Wandel (pp. 105-112). Bielefeld: WBV. Schramm, Karen (2016). Empirische Forschung. In Daniela Caspari; Friederike Klippel, Michael K. Legutke; Karen Schramm (Eds.), Forschungsmethoden in der Fremdspra‐ chendidaktik - Ein Handbuch (pp. 49-58). Tübingen: Narr. Schramm, Stefanie (2010). Spracherwerb - Was können wir noch lernen? Zeit Wissen. Retrieved from http: / / www.zeit.de/ zeit-wissen/ 2010/ 06/ fremdsprache-lernen-alter Schreier, Margrit (2012). Qualitative Content Analysis in Practice. Thousand Oaks: Sage. Schreier, Margrit (2014a). Varianten qualitativer Inhaltsanalyse: Ein Wegweiser im Di‐ ckicht der Begrifflichkeiten. Forum Qualitative Sozialforschung/ Forum: Qualitative So‐ cial Research, 15(1). Retrieved from http: / / nbn-resolving.de/ urn: nbn: de: 0114-fqs1401185 Schreier, Margrit (2014b). Qualitative Content Analysis. In Uwe Flick (Ed.), The SAGE Handbook of Qualitative Data Analysis (pp. 170-183). London: SAGE. Settersten, Richard A. Jr.; Angel, Jacqueline L. (2011). Trends in the Sociology of Aging: Thirty Year Observations. In Richard A. Jr Settersten; Jacqueline L. Angel (Eds.), Handbook of Sociology of Aging (pp. 3-13). New York: Springer. Settinieri, Julia (2016). Deskriptiv- und Inferenzstatistik. In Daniela Caspari; Friederike Klippel; Michael K. Legutke; Karen Schramm (Eds.), Forschungsmethoden in der Fremd‐ sprachendidaktik - Ein Handbuch (pp. 324-340). Tübingen: Narr. Shakespeare, William (1996). The Tempest. Frank Günther (Ed.). (Original work published 1623). München: dtv. Shavelson, Richard J.; Hubner, Judith T.; Stanton, George C. (1976). Self-Concept: Vali‐ dation of Construct Interpretations. Review of Educational Research, 46(3), 407-441. Singleton, David; Ryan, Lisa (2004). Language Acquisition: The Age Factor (2nd ed.). Cle‐ vedon: Multilingual Matters. 251 Works Cited Skaalvik, Einar M.; Skaalvik, Sidsel (2008). Teacher Self-Efficacy. Conceptual Analysis and Relations with Teacher Burnout and Perceived School Context. In Dennis M. McInerney; Rhonda G. Craven; Herbert W. Marsh (Eds.), Self-processes, Learning, and Enabling Human Potential: Dynamic New Approaches (pp. 223-247): Information Age Publishing. Smith, Jacqui; Ryan, Lindsay H. (2015). Psychological Vitality in the Oldest Old. In K. Warner Schaie (Ed.), Handbook of the Psychology of Aging (pp. 303-319). London: El‐ sevier. Staehelin, Hannes B. (2005). Promoting Health and Wellbeing in Later Life. In Johnson L. Malcolm (Ed.), The Cambridge Handbook of Age and Ageing (pp. 165-177). Cam‐ bridge: CUP. Staudinger, Ursula M.; Heidemeier, Heike (2009). Bildungspsychologie des höheren Er‐ wachsenenalters. In Christiane Spiel; Barbara Schober; Petra Wagner; Ralph Reimann (Eds.), Bildungspsychologie (pp. 192-209). Göttingen: Hogrefe. Swann, William B. J.; Bosson, Jennifer K. (2010). Self and Identity. In Susan T. Fiske; Daniel T. Gilbert; Gardner Lindzey (Eds.), Handbook of Social Psychology (pp. 589-628): Wiley. Szasz, Thomas S. (1973). The Second Sin. New York: Anchor Press/ Doubleday. Tashakkori, Abbas; Teddlie, Charles (Eds.). (2010). Sage Handbook of Mixed Methods in Social and Behavioral Research. (2nd ed.). Thousand Oaks: Sage. Taylor, Florentina (2011). A Quadripolar Model of Identity in Adolescent Foreign Language Learners. (PhD Thesis), University of Nottingham. Taylor, Florentina (2014). Relational Views of the Self in SLA. In Sarah Mercer; Marion Williams (Eds.), Multiple Perspectives on the Self in SLA (pp. 92-108). Bristol: Multilin‐ gual Matters. Teddlie, Charles; Tashakkori, Abbas (2008). Quality Inferences in Mixed Methods Re‐ search: Calling for an Integrative Framework. In Manfred M. Bergman (Ed.), Advances in Mixed Methods Research (pp. 101-119). Thousand Oaks: Sage. Teddlie, Charles; Tashakkori, Abbas (2010). Overview of Contemporary Issues in Mixed Methods Research. In Abbas Tashakkori; Charles Teddlie (Eds.), Sage Handbook of Mixed Methods in Social and Behavioral Research (2nd ed., pp. 1-41). Thousand Oaks: Sage. Tippelt, Rudolf; Schmidt, Bernhard; Schnurr, Simone; Sinner, Simone; Theisen, Catharina (2009). Bildung Älterer - Chancen im demografischen Wandel. Bielefeld: WBV. 252 Works Cited Tomasello, Michael (1993). On the Interpersonal Origins of Self-concept. In Ulric Neisser (Ed.), The Perceived Self (pp. 174-184). New York: CUP. Turner, Julianne C.; Meyer, Debra K. (2000). Studying and Understanding the Instruc‐ tional Contexts of Classrooms: Using our Past to Forge the Future. Educational Psy‐ chologist, 35(2), 69-85. Turner, John C.; Onorato, Rina S. (1999). Social Identity, Personality, and the Self-Concept: A Self-Categorization Perspective. In Tyler, Tom R. (Ed.), The Psychology of the Social Self. (pp. 11-46). New Jersey: Erlbaum. Ushioda, Ema (2014). Motivational Perspectives on the Self in SLA: A Developmental View. In Sarah Mercer; Marion Williams (Eds.), Multiple Perspectives on the Self in SLA (pp. 127-141). Bristol: Multilingual Matters. Ushioda, Ema; Dörnyei, Zoltán (2009). Motivation, Language Identities and the L2 Self: A Theoretical Overview. In Zoltán Dörnyei; Ema Ushioda (Eds.), Motivation, Language Identity and the L2 Self (pp. 1-8). Bristol: Multilingual Matters. Victor, Christina (2010). The Demography of Ageing. In Dale Dannefer; Chris Phillipson (Eds.), The SAGE Handbook of Social Gerontology (pp. 61-74). Thousand Oaks: SAGE. Völkening, Gertrud (2006). Demographischer Wandel als Herausforderung für kommu‐ nale Bildungseinrichtungen. REPORT, 29, 51-59. Walker, Alan (2010). Ageing and Quality of Life in Europe. In Dale Dannefer; Chris Phil‐ lipson (Eds.), The SAGE Handbook of Social Gerontology (pp. 573-586). Thousand Oaks: Sage. Wang, Mo; Shi, Junqi (2015). Work, Retirement and Aging. In K. Warner Schaie; Sherry L. Willis (Eds.), Handbook of the Psychology of Aging (8th ed., pp. 339-359). London: Elsevier. Waxenegger, Andrea (2011). Lernen und Bildung im späteren Lebensalter. Leitlinien und Prioritäten 2020. Graz: Thalerhof. Weil, Joyce (2015). Applying Research Methods to a Gerontological Population: Matching Data Collection to Characteristics of Older Persons. Educational Gerontology, 41(10), 723-742. Weiland, Severin; Wittrock, Philipp (2014). Thierse-Reaktion auf geheime Zitate: "Kohls Wut scheint etwas sehr Stabiles zu sein". Retrieved from www.spiegel.de/ politik/ deutschland/ helmut-kohl-spd-politiker-thierse-kritisiert-altkanzler-a-995620.html 253 Works Cited Wiarda, Jan-Martin (2010). Wissen sammeln um die Ecke. Die Zeit. Retrieved from www.zeit.de/ 2010/ 48/ C-Volkshochschule-Erwachsenenbildung Withnall, Alexandra (2012). Lifelong or Longlife? Learning in the Later Years. In David N. Aspin; Judith Chapman; Karen Evans; Richard Bagnall (Eds.), Second International Handbook of Lifelong Learning - Part One (pp. 649-664). London: Springer. Xu, Man K.; Marsh, Herbert W.; Hau, Kit-Tai; Ho, Irene T.; Morin, Alexandre J. S.; Ab‐ duljabbar, Adel S. (2013). The Internal/ External Frame of Reference of Academic Self-concept: Extension to a Foreign Language and the Role of Language of Instruction. Journal of Educational psychology, 105(2), 489-503. 254 Works Cited Appendix A: Questionnaire 256 Appendix A: Questionnaire 257 Appendix A: Questionnaire 258 Appendix A: Questionnaire Appendix B: Letters to Heads of Language Departments Appendix C: Letter to vhs Centre Language Instructors Appendix D: Interview Questions Introduction (preamble) Information about my Study (learners of English at vhs centre) Information about data protection, participant codes, use/ storage of recorded interviews Inform participants about the interview procedure (e.g. if sth. is unclear) Questions on his/ her behalf? o Topics/ Questions + probes & prompts  o Notes/ Comments Warm-Up As I have mentioned earlier, I am interested in what makes someone go and attend English classes at the local adult education centre (vhs)? Why have you decided to take an English class at the vhs? Probes: When exactly was this? What was it like for you to be back in „school“ in the beginning? (especially for former teachers) How did people in your environment (family and friends) react to your decision? How do you motivate yourself to stay on track and keep on learning? What do you think influences your motivation in a good/ bad way? What expectations do you have about vhs classes (compared to e.g. English at school)?  o  o  o  o  o  o  o Main part I gave you the opportunity to do a reflection task on your foreign language learning biography. If you‘d like to we could take a look at how you answered the questions and at how you drew your graph. Could you explain your story and graph to me a little? Prompts: Could you give me more information about your graph? Why does the line go like this...? At this point is a decline/ increase/ high/ low why is that? What happened at that time? In case of a low: what helped you to get over it and stay motivated? Why exactly have you picked these three incidents? Could you tell me a little bit more about them? Now after looking at your graph, would you make any changes? (add/ leave out) Probes: o Working life/ retirement: changes in learning English at vhs?  o  o  o  o  o  o Do you remember a very nice/ horrible English lesson? Prompts:  Was there a situation in class that you remember to be very positive/ negative?  Was there a task or a text etc. that you found to be odd or strange at first?  Was there a situation in an English lesson that annoyed or amused you? Probes: o How did you feel during this situation or afterwards? o What was on your mind? o How did you handle/ think about the situation afterwards?  o  o  o  o  o  o  o When you look at your English learning experiences in the past and you compare them to your situation/ level today: Have you changed your ways of learning English? Prompts:  Do/ Did you have some kind of routines when learning English?  How do you approach the task of learning English? Do you have certain strategies/ methods?  o  o  o 262 Appendix D: Interview Questions In your opinion, what kind of characteristic does someone who would like to learn English need in class? Prompts: What makes a good/ successful English class participant? Which character traits should he/ she have? Probes: Which of these characeristics do you possess? Do you have „idols“ in class or other participants you look up to/ admire? Which characteristics would you like to possess (more) or which characteristics do you think would you need (would be useful) to be better at learning English right now? How do you know that you are missing/ lacking this characteristic?  o  o  o  o  o How would you describe the following things in three words (e.g. adjectives) : yourself as a person yourself as a foreign language learner your learning of English  o Which role do you play in your class? (team, contributions...)  o What role does the teacher play when you learn English? Probes: What do you expect from an English teacher? Or: what makes a ideal teacher for you? How important is the teacher to you? What do you think about feedback/ corrections in class? What do you think: How happy were your teachers in the past/ is your teacher with your English skills?  o  o  o  o  o How happy are you with your English skills (level) right now? Prompts: on a scale from 1 to 10? Probes: How confident do you feel a) in class b) abroad? How about in terms of different language skills? (speaking, reading, listening, writing, vocab, grammar...? )  o  o  o  o 263 Appendix D: Interview Questions How do you picture your further learning/ future time in your English class? Probes: What are your goals concerning your English skills for the next months/ years/ the future? Would you like to change your ways of learning English? (strategies, methods...) What would you like to do more often/ more intensely inand outside your English class? Is there anything you could do without when learning English in- / outside English class?  o  o  o  o  o What advice would you give someone who is about to start learning English/ taking up an English class? Prompts: What would he/ she need to feel confident/ comfortable in class or to fit in? Probes: What kind of advice would you give him/ her in more general terms of language learning? (strategies, motivation, procedure, classes, teacher...)  o  o  o Closing Would you like to add anything? Is there anything you would like to say that I haven‘t asked you? Do you have any questions for me?  o  o Thank you very much for your time and the very interesting interview! Notes after the interview: 264 Appendix D: Interview Questions Appendix E: Informed Consent - Interviews Appendix F: Sample Learning Biographies Adam (G3TM01) Ulf (G1TM03) 267 Appendix F: Sample Learning Biographies Wolfram (G1TM04) 268 Appendix F: Sample Learning Biographies Appendix G: Transcription Rules (based on: Dresing et al. 2015: 27-32) Transcription Meaning / Translation I Interviewer (..), (4) Pauses; each “.” indicates one second; if a pause is longer than three seconds the number is inserted in brackets. […] [transl. lit.: lazy dog] [laughs] Shortened passages; comments or literal translations; ac‐ tions / Incomplete sentences or discontinuations / / Speech overlap REAlly Capitalized letters indicate special emphasis in speech (e.g. intonation or volume) hedgehog English in the original interview Appendix H: Interview Summaries Participant (pseudonym, code, age, re‐ tired/ not retired, year they joined current course) Interview Summary Barbara (G1TF01) 59 NR 1982 I loved English as a subject at school and always had the ambition to get high grades and always did my homework, which was then passed around in class and copied by my classmates. After school, however, I quickly realized that I wasn’t prepared for conversations in English with native speakers at work or for work purposes. I still have that am‐ bition and passion for learning English. I travel to the U.S. regularly and love to speak to the people there. Peter (G1TM01) 65 R 1982 I became really interested in learning English at school. I wanted to be able to understand and play the songs by bands such as the Rolling Stones, the Beatles etc. which were very popular back then. I listened to them on pirate radio stations and was playing in a band. Luckily our English teacher at school gave us some good foundations - even though it was just “school English”. After school I did not do much with my English skills but soon realized when we went on our first holidays to the United States how much I had forgotten or how little that English from school had to do with the English in the real world. We then decided to take up this course in the 1980s and even though there was a break in-between vhs courses, I can say that I have become more confident on our holidays. But I am still not sure about my pronunciation or if others really understand me. I also need quite some time to formulate sentences in English. Astrid (G1TF02) 52 NR 2012 I come from a country, where English is more common in everyday life than in Germany. I moved to Germany 18 years ago and have taken German classes at the vhs centre earlier. After an English intensive course abroad, I wanted to main‐ tain my English skills to be prepared for occasional trips abroad and decided to join this English class. Participant (pseudonym, code, age, re‐ tired/ not retired, year they joined current course) Interview Summary Anne (G1TF03) 67 R 2013 I joined this class together with my neighbour and very good friend only recently. My daughter-in-law has also been in this class for a while and recommended it to us. But it is also the reason why I had doubts for a long time at the beginning if I am a genuinely accepted member of the group or just “the mother-in-law of…”. I have had the feeling recently that I’ve been accepted. Generally, I love to travel to Africa and soon Canada and I also would like to do something to keep my mind active, too. However, I am very - or in fact too ambi‐ tious about it so that sometimes when the pressure I put on myself gets too big, I would rather stay at home than go to class in danger of performing below my or other’s expecta‐ tions. Dorothea (G1TF04) 61 R 2013 I’m curious about a lot of things and keep myself very busy. I think broadening one’s mind or perspective is very im‐ portant to me. I do occasionally need English when I visit my son, who is living and working in Dubai. But other than that, I don’t have many opportunities to use my English skills. The vhs class gives me that opportunity. But my French skills are better than my English ones. I tried to take up Italian once, but that was too much: three languages. I simply don’t have the time or ambition. I wish I had the ambition to sit down at home and go over the textbook con‐ tents again. But I am actually pretty relaxed about this si‐ tuation - it’s okay like this. Christine (G1TF05) 63 R 1986 After getting my hunting license in the 80s, noticed that I wanted to keep learning and decided to take up an English class at the vhs centre. I also hoped this would help me sup‐ port my young son with his English homework - but that soon turned out to be too much to expect from such a course. I can say that I mainly attend this class because of the group or the people in the group. However, with my current tea‐ cher leaving and since I don’t like the new teacher, I’m thin‐ king about leaving the class. I will still attend another Eng‐ lish class in another town. I don’t expect too much from this course or myself, simply to avoid the feeling of stress or pressure. I do it ‘just for fun’. I think my English has dete‐ riorated a little since my husband is not dealing with cus‐ tomers from England anymore. 271 Appendix H: Interview Summaries Participant (pseudonym, code, age, re‐ tired/ not retired, year they joined current course) Interview Summary Christian (G1TM02) 64 R 1992 I started this English course to be prepared for our first bigger holidays in countries such as New Zealand, Canada, Thailand etc. but also did an intensive course in Torquay in the 90s, because I needed it for work then. Because I needed it for my job quite often to communicate with business partners in Europe, my English was quite good. But I quickly lost most of my English skills and I think that after my 50s my memory or English skills also deteriorated quickly. It became increasingly difficult to process things in class and to memorize new words. I just want to learn normal English - because the English you learn in class or read in the book is not like the English they speak in England. Ulf (G1TM03) 61 NR 1986 Travelling the States and English country music have mo‐ tivated me to take up this course in the 80s. Our teacher was good in the beginning but got sick and his teaching perfor‐ mance was impacted by it as well. That also impacted my motivation, but also the fact that I was working long hours. The change of teachers and different teaching style picked me up a little and I was forced to participate or speak more often in class. Of course I wanted it as well. But what is also important to me is that there is no pressure to do extra work or in forms of tests. It would really upset me if a new teacher would introduce such things. I don’t know yet if I will con‐ tinue with the course since I am self-employed and busy with my grandchild, too. I’m often too tired in the evening to go to class. I will see how I get along with the new teacher and then decide. Wolfram (G1TM04) 70 R 1994 I went to a grammar school, where English was my third foreign language - after Latin and Greek. Due to my other language learning experiences, I feel quite comfortable with the grammar. However, I find my vocabulary is still limited. I have noticed those deficits when talking to other resear‐ chers at international conferences and felt quite inhibited as I feel that you are being judged based on your English skills at these occasions. I’m a perfectionist - not only when it comes to using English - and thus, need some time to pre‐ pare a correct sentence while the conversation in English has already progressed. I would like to be able to simply go ahead and speak up more freely or carelessly as some other learners in my class. Nevertheless, I am a realist and don’t expect any big improvement in that area as well as in terms of an expansion of my vocabulary. 272 Appendix H: Interview Summaries Participant (pseudonym, code, age, re‐ tired/ not retired, year they joined current course) Interview Summary Diana (G2TF01) 73 R 2002 I have started to learn English when I was 30 at a Volks‐ hochschule (’74 until ’80). I enjoyed it very much and even took up French during that time after having learned some English for four years. I have profited a lot from reading books in English at home. Once I retired I joined this con‐ versation class - I noticed when I was travelling with my husband that my English is not as fluent as it used to or should be. I also wanted to do something for my brain. Even though I struggled to keep up at the beginning, I am now much more confident but still not perfect. Alexandra (G2TF02) 79 R 2001 I used to be an English teacher for high school children. Three years into my retirement I noticed how my English skills declined. Since I and my husband have a passion for languages, I decided to join this conversation class to main‐ tain my English skills. Personally, I actually prefer French a little bit over English, but there are no classes for my level in the area. I know that my English vocabulary lacks more “modern” words, but I am nevertheless happy with what I have and don’t see a need to learn them. Because I may have more experience with the English language, I don’t want to stand out or impose my preferences (e.g. re‐ ading poems) on the others in class. Maria (G2TF03) 65 R 1986 At the beginning of my vhs experience, me and another course participant decided to study for an English certifi‐ cate. I think that was one of my major highlights during all this time at vhs because I had a goal and my skills were quite good compared to now. Sometimes it is hard to get off the couch in the evening and go to class. But the group itself draws me back and motivates me. Unlike learning English at school, my English classes at vhs have taught me more how to speak it - after all, this is what I need when I travel. Brigitte (G2TF04) 65 R ca. 1988 I was not particularly good or interested in languages at school. And if a teacher wanted to introduce homework or tests, I would probably leave since I don’t want this pressure anymore. I decided to take this course to maintain or refresh my English skills and keep my mind active. I also decided early on to do an English certificate. Passing the test for this certificate gave me a feeling of accomplishment. I feel con‐ nected to the group which is why I don’t need to motivate myself greatly to come here. I wish there were more people in our class. 273 Appendix H: Interview Summaries Participant (pseudonym, code, age, re‐ tired/ not retired, year they joined current course) Interview Summary Jennifer (G2TF05) 68 R 2013 I joined this course to prepare myself for a short trip to Ire‐ land. The atmosphere in a group like this - how the other learners and the teacher behave and interact - is very im‐ portant to me. My first experiences with learning English at school and while I was working were rather boring, dry and associated with fear. We didn’t learn to speak or listen to recordings. A few bad experiences in classes when I could not say anything have put me down. It wasn’t until my job as a primary school teacher required me to also teach Eng‐ lish to kids and I went on an intensive course in England, that I realized I needed to go to the country where the target language is spoken. I need to hear the language in the country and look up words in a monolingual dictionary to get the most out of it. I think over the years I have become more self-confident and learned to speak my mind - also in class, if I don’t like something. Lena (G2TF06) 69 R 2008 After I retired from my job as a maths and physics teacher a few years ago, I joined this course to improve my skills for holidays abroad. I noticed that I have enjoyed using it on holidays in the States or Britain. I am not particularly ta‐ lented when it comes to learning languages but I have al‐ ways been curious about it and I’m interested in the British culture and people. I have stayed in Britain as an au pair and later lived there with my husband during his studies. I have noticed on more recent travels to English-speaking count‐ ries that the course at vhs does help to overcome the inhi‐ bition to speak. This is why the speaking component of this English course is so important to me. Tanja (G3TF01) 72 R 2003 I started with English at school and luckily my parents had a pub frequented by American soldiers that were stationed nearby. This way, I could already use some of the English I learned back then. My daughter lives with her family in the United States. Me and my husband have visited them several times for a couple of weeks. She wants us to speak German with our grandchildren. However, I do learn the occasional English words from my grandchildren, too, and I can con‐ verse with my daughter’s late mother-in-law, my son-in-law and other people in their environment. I sometimes thought about joining another English group that is closer to where I live and on a higher level, but I think I have connected very well with this group already and I’m afraid we lack the mi‐ nimum appropriate number of participants without me. I try to learn new words at home by posting them on cupboards in my kitchen and going over them while I’m cooking. 274 Appendix H: Interview Summaries Participant (pseudonym, code, age, re‐ tired/ not retired, year they joined current course) Interview Summary Katharina (G3TF04) 67 R 2008 I first started to learn English while working as a nurse in my 20s. I needed English skills to get my degree at an eve‐ ning school. I joined this English class shortly after retiring since I now had much more time and was approaching this in a more relaxed way. My English has never been good and I always had to work very hard for the skills I have now - I was never that talented as other people who could just “talk away” in classes. I still struggle to speak it but I do know the grammar quite well because I had a good tutor back then when I started to learn English. Later on while I was still working I did try some English courses, I learned some French and a little bit of Italian. I enjoy learning English and being in our group. I want to be on the same level with everyone. Not above them, and certainly not the worst in class - that would intimidate me. Ruth (G3TF06) 64 R 2012 I recently retired from my job as a primary school teacher and was looking for more hobbies I could do in my free-time. I find it useful for travelling and keeping my mind active to learn languages. Even though I only recently joined this class, I do notice that I have progressed with my English since I feel words coming back and I feel less inhibited to speak it. Katja (G3TF07) 75 R 2013 Before this course, I was in another course but was very frustrated with the way things went there. I hardly had the chance to speak because either the teacher or only one or two other learners were talking the whole time. It is much better in this class and I enjoy learning in the group, espe‐ cially when we can talk to a partner or a small group work. As a former primary school teacher, I enjoy communicating with people and I’m happy that the group has accepted me so well. I actually like learning English at school but some teachers back then did not make it enjoyable for me and I think that my generation was generally more inhibited than today’s youth. Going on holidays is the most wonderful thing for me and I love to make contact with people there, which is why it is so important for me to improve my spea‐ king skills and become less inhibited. However, I don’t feel that inhibited anymore when I need to speak English abroad - this has definitely changed. 275 Appendix H: Interview Summaries Participant (pseudonym, code, age, re‐ tired/ not retired, year they joined current course) Interview Summary Birgit (G3TF09) 62 R ca. 2009 After retiring from my job as a teacher, I was looking for something where I could keep my mind active. Learning English is very useful if I want to travel. However, I think in order to be a more successful learner I need more oppor‐ tunities to use it outside the classroom. But how am I sup‐ posed to do this? Even if I travel abroad but in a group of tourists, they usually speak in German. If the topic is inte‐ resting or useful for me and if I have a textbook I can refer to, I feel quite comfortable in class. Adam (G3TM01) 81 R ca. 2004 My main reasons for coming to class have not so much to do with learning the language itself (I don’t need English anymore since I stopped travelling). It is more about pre‐ venting my memory from declining even more (I have no‐ ticed it getting worse every day, although I am still in a better shape than others in my age) and also to get out of my house and spend some time with the others in class. It is a really nice group and the highlight of every semester is our end-of-semester lunch. I have also taken some computer classes here at this vhs centre, but haven’t been very suc‐ cessful with applying this at home. I don’t know if I will continue this class after this semester as my pension is very small and I cannot get a discount from the vhs centre. The long drive to the centre by car has become increasingly dif‐ ficult for me. 276 Appendix H: Interview Summaries Abbildungsverzeichnis Figure 3.1 Facet analysis of self-concept by Hattie (1992: 37 - bold by me) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Figure 4.1 My MMR Research Design based on the Parallel Convergent Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Figure 4.2 Convergent Parallel Design (Creswell & Plano Clark 2011: 69) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 Figure 5.1 Sample Item for Ideal L2 Self . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Figure 5.2 Likert Scale example illustrating the values of 1 to 6 with the sample sentence "I like to drink tea" (Ich trinke gerne Tee). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 Figure 5.3 Adjective Scale in Question 3a . . . . . . . . . . . . . . . . . . . . . . . . 112 Figure 5.4 Motives for Learning English . . . . . . . . . . . . . . . . . . . . . . . . . 120 Figure 5.5 Adjective Scale: Active-Passive . . . . . . . . . . . . . . . . . . . . . . . . 125 Figure 5.6 Adjective Scale: Distracted-Focused . . . . . . . . . . . . . . . . . . . . 126 Figure 5.7 Adjective Scale: Inhibited-Relaxed . . . . . . . . . . . . . . . . . . . . . 127 Figure 5.8 Adjective Scale: Open-Careful Towards New Things . . . . . 128 Figure 5.9 Adjective Scale: Strong-Weak (NR/ R) . . . . . . . . . . . . . . . . . . . 129 Figure 5.10 Adjective Scale: Strong-Weak (u60/ 60+) . . . . . . . . . . . . . . . . 129 Figure 5.11 Adjective Scale: Cooperative-Independent (u60/ 60+) . . . . . 130 Figure 5.12 Adjective Scale: Cooperative-Independent (M/ F) . . . . . . . . . 131 Figure 5.13 Adjective Scale: Content-Ambitious . . . . . . . . . . . . . . . . . . . 132 Figure 5.14 Adjective Scale: Introverted-Extroverted . . . . . . . . . . . . . . . 133 Figure 7.1 MMR Research Design (Final Merging Stage highlighted in green) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 Tabellenverzeichnis Table 2.1 Chronological Age Spans for the Young-Old . . . . . . . . . . . . . . . 25 Table 3.1 Self-State Representations (after Higgins 1987) . . . . . . . . . . . . . 59 Table 5.1 Likert Scale Items and Operationalized L2-Constructs . . . . . . . 105 Table 5.2 Survey Participants: Age Spans . . . . . . . . . . . . . . . . . . . . . . . . . . 117 Table 5.3 Survey Participants: Working and Retired . . . . . . . . . . . . . . . . . 117 Table 5.4 Survey Participants: Time spent at vhs Centre . . . . . . . . . . . . . 118 Table 5.5 Survey Participants: Male and Female . . . . . . . . . . . . . . . . . . . . . 119 Table 6.1 Interview Questions and their Theoretical Target(s). . . . . . . . . 144 Table 6.2 Interview Participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151 Table 6.3 Category Formation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 Table 6.4 Category System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 Table 6.5 Overview of L2-Vision Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 Table 6.6 Examples for L2-Vision Sub-Categories . . . . . . . . . . . . . . . . . . . 165 Table 6.7 L2-Vision Groups (Time at vhs Centre) . . . . . . . . . . . . . . . . . . . . 166 Table 7.1 Merging Results: Ambition and Intended Effort . . . . . . . . . . . . 205 Table 7.2 Merging Results: Resilience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 Table 7.3 Merging Results: Critical Incidents Abroad and L2-Vision . . . . 209 Table 7.4 Merging Results: Confidence: Taking it Easy . . . . . . . . . . . . . . . 211 Table 7.5 Merging Results: Defying the Effects of Ageing . . . . . . . . . . . . 213 Table 7.6 Merging Results: Now and Then - Ways of Learning a Language . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215 Table 7.7 Merging Results: Now and Then - Learning Context . . . . . . . . 216 Table 7.8 Merging Results: Perceived Language Learning Progress . . . . 218 Table 7.9 Merging Results: Language Learning and the Social Sphere . . 220 Giessener Beiträge zur Fremdsprachendidaktik Young-old learners are an underresearched group in foreign/ second language research. The present mixed-methods study aims to provide a more differentiated view of this group in the context of lifelong learning and, more specifically, learning English as a foreign language. The author draws from concepts in gerontology, psychology, adult education, and foreign/ second language research to investigate the L2-selfconcepts of young-old language learners at Volkshochschulen in Germany. ISBN 978-3-8233-8311-6