UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A case study of the BalancedView course : addressing weight stigma among health care providers in British… O'Reilly, Caitlin Joyce 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2018_may_o'reilly_caitlin.pdf [ 3.22MB ]
Metadata
JSON: 24-1.0364065.json
JSON-LD: 24-1.0364065-ld.json
RDF/XML (Pretty): 24-1.0364065-rdf.xml
RDF/JSON: 24-1.0364065-rdf.json
Turtle: 24-1.0364065-turtle.txt
N-Triples: 24-1.0364065-rdf-ntriples.txt
Original Record: 24-1.0364065-source.json
Full Text
24-1.0364065-fulltext.txt
Citation
24-1.0364065.ris

Full Text

A CASE STUDY OF THE BALANCEDVIEW COURSE: ADDRESSING WEIGHT STIGMA AMONG HEALTH CARE PROVIDERS IN BRITISH COLUMBIA by  Caitlin Joyce O’Reilly  B.S.W., The University of Victoria, 2008 M.P.P., Simon Fraser University, 2011  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Kinesiology)   THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) February 2018 © Caitlin Joyce O’Reilly, 2018 ii  Abstract Growing evidence shows weight stigma as a problem in health care settings. However, there remains a lack of conceptual clarity – particularly regarding if and how the medicalization of weight is implicated in weight stigma – and a gap in knowledge about how to successfully reduce weight stigma in health care. The research questions that guided this study were thus: • What are the different ways that weight stigma in health care can be conceptualized? • What strategies can be employed to reduce weight stigma among health care providers? These questions were explored through a mixed methods case study of the development and implementation of an online course on weight stigma for health care providers in British Columbia called BalancedView, sponsored by the Provincial Health Services Authority. Using participant observation, document analysis, a focus group and semi-structured interviews, I examined how health care stakeholders who developed the course, and participants who went on to take the course, conceptualized weight stigma. I evaluated the effects of the course on 249 participating health care providers through questionnaires before and after the course. Using interviews with course participants and documentary analysis of qualitative comments made by participants during the course, I also explored what was most helpful about the course and why. Following a thematic analysis, I show how weight stigma was conceptualized as a process involving biased attitudes and beliefs that lead to discriminatory behaviours and adverse outcomes. It was perceived as a causally complex issue, with a relationship to emotions. The extent to which the medicalization of weight was viewed as part of weight stigma was a divisive topic in the development stage of the course. However, many participants who took the course reflected later that after learning about medicalization they saw harms in medicalized approaches to weight in health care.  iii  This study contributes to the currently limited literature on weight stigma reduction in health care. I demonstrate how an online course on weight stigma that uses multiple stigma reduction techniques had a positive effect in terms of reducing participants’ weight bias and discuss what essential elements within such interventions should be.  iv  Lay summary  Growing evidence demonstrates that weight stigma is a problem in health care settings. The aims of this work were: 1) to better understand the concept of weight stigma in health care; and 2) to explore how to reduce it. These aims were investigated through a case study of the development and implementation of an online course on weight stigma for health care providers in British Columbia called BalancedView. I examined how stakeholders who developed the course and participants who took the course conceptualized weight stigma. Among the 249 health care providers who took the course I evaluated its effects and looked at what was most helpful about the course and why. Through this study I was able to build understandings of weight stigma and show how an online course on weight stigma that uses multiple stigma reduction techniques can have a positive effect in terms of reducing participants’ weight bias.  v  Preface This project was a partnership project undertaken with British Columbia (BC) Mental Health and Substance Use Services, an agency of the Provincial Health Services Authority (PHSA) in BC. Through this partnership, an online course on weight bias and stigma was developed for health care providers called BalancedView, as led by the PHSA. I was one of many individuals involved in helping develop this course alongside PHSA staff and other collaborators. Collaborators included members of a committee of health care providers in BC, members of a committee of experts in weight stigma and independent contractors (i.e. a curriculum development consultant, two scoping review consultants and an evaluation consultant). My role in the course development was to participate with the committees in advising on content and format.   The data collection tools used in my study were also developed through collaboration with the PHSA, the committees described above and the evaluation consultant. The evaluation consultant and myself, in partnership with PHSA staff and the committees, developed the questionnaire and focus group guide. I developed the interview guide with input from the above parties.  I conducted participant observation and interviews. I co-facilitated the focus group with the evaluation consultant. Longitudinal follow-up questionnaire data were analyzed in partnership with the evaluation consultant (n = 74). I analyzed the qualitative and quantitative data collected through the pre-course and post-course questionnaires for the larger sample (N = 249), as well as the interview and focus group data and scripted reflections participants engaged in within the course. vi  A portion of the findings from chapters four and five has been published: O’Reilly, C. (2016). Considerations in mitigating weight stigma through health professional education. In E. Cameron & C. Russell (Eds.). The fat pedagogy reader: Challenging weight-based oppression in education. New York, NY: Peter Lang Publishing. The literature review conducted for my dissertation also informed the following publication: Cameron, E., & O'Reilly, C. (2015). Type 2 diabetes in youth. Biochemistry and Cell Biology, 93(5), 430. Dr. Cameron and myself contributed equally to this review article.  This study was reviewed by the Behavioural Research Ethics Board at the University of British Columbia (certificate # H13-02944). vii  Table of contents  Abstract .......................................................................................................................................... ii	Lay summary ................................................................................................................................ iv	Preface .............................................................................................................................................v	Table of contents ......................................................................................................................... vii	List of tables................................................................................................................................ xiii	List of figures .............................................................................................................................. xiv	List of text boxes ........................................................................................................................... xv	List of abbreviations ................................................................................................................. xvii	Acknowledgements .................................................................................................................. xviii	Dedication ..................................................................................................................................... xx	Chapter 1: Introduction ................................................................................................................1	Chapter 2: Literature review ........................................................................................................7	2.1	 How is stigma conceptualized in the research literature? .................................................. 7	2.2	 How is weight stigma conceptualized in the research literature? .................................... 10	2.2.1	 Psychosocial approaches to weight stigma ............................................................... 11	2.2.2	 Other approaches to weight stigma ........................................................................... 15	2.3	 Prevalence and implications of weight stigma ................................................................. 26	2.3.1	 Weight stigma in the health sector ............................................................................ 27	2.4	 What does current research say about weight stigma reduction? .................................... 31	2.4.1	 Attribution theory and weight stigma reduction ....................................................... 33	2.4.2	 Social consensus theory and weight stigma reduction .............................................. 38	viii  2.4.3	 Evoking empathy and weight stigma reduction ........................................................ 39	2.4.4	 Contact theory approaches to weight stigma reduction ............................................ 45	2.4.5	 Cognitive dissonance or consciousness raising ........................................................ 46	2.4.6	 Multi-method interventions ...................................................................................... 48	2.4.7	 Health At Every Size and its potential for stigma reduction .................................... 52	2.5	 Other insights into weight stigma reduction .................................................................... 53	2.6	 Gaps in the literature: What do we need to know more about? ....................................... 57	Chapter 3: Methodology ..............................................................................................................60	3.1	 Research questions ........................................................................................................... 60	3.2	 A case study research approach ....................................................................................... 61	3.3	 Informed by a pragmatic research approach .................................................................... 65	3.4	 The case of BalancedView ............................................................................................... 67	3.4.1	 Development description .......................................................................................... 68	3.4.2	 Pilot test description .................................................................................................. 73	3.4.3	 Launch description .................................................................................................... 74	3.4.4	 The course content .................................................................................................... 75	3.5	 Data collection methods ................................................................................................... 77	3.5.1	 Participant observation, reflexivity and research journal ......................................... 81	3.5.2	 Documents as sources of data ................................................................................... 82	3.5.3	 Questionnaires ........................................................................................................... 83	3.5.4	 Semi-structured interviews with course participants ................................................ 90	3.5.5	 Group interview and focus group with committee members .................................... 92	3.6	 Recruitment ...................................................................................................................... 94	ix  3.7	 Sample .............................................................................................................................. 96	3.8	 Analysis.......................................................................................................................... 100	3.8.1	 Reflexivity ............................................................................................................... 103	3.9	 Ethics .............................................................................................................................. 105	Chapter 4: Findings regarding research question one / Conceptualizing weight stigma....109	4.1	 How was weight stigma conceptualized during the development of BalancedView? .. 109	4.1.1	 A story of shifting conceptualizations in BalancedView’s development – an interlude .............................................................................................................................. 109	4.1.2	 Stigma as process .................................................................................................... 119	4.1.3	 Stigma as negative attitude/belief/stereotypes ........................................................ 120	4.1.4	 Stigma as discriminatory behaviour or outcome .................................................... 122	4.1.5	 Stigma as causally complex .................................................................................... 126	4.1.6	 Medicalization as stigmatizing… or not?: A divisive issue .................................... 130	4.1.7	 Stigma as related to emotions: Emotions matter to how weight stigma is conceptualized ..................................................................................................................... 134	4.2	 How was weight stigma conceptualized by participants during the implementation of BalancedView? ....................................................................................................................... 138	4.2.1	 Weight stigma as a process involving attitudes/beliefs and actions/unfair  treatment ............................................................................................................................. 139	4.2.2	 Controllability beliefs as stigmatizing .................................................................... 140	4.2.3	 Disease beliefs: Medicalization as more harmful than helpful? ............................. 141	4.2.3.1	 Medicalization as harmful and stigmatizing .................................................... 142	4.2.3.2	 Pros and cons to medicalization ....................................................................... 147	x  4.2.3.3	 Medicalization as positive and needed ............................................................ 150	Chapter 5: Findings regarding research question two / Reducing weight stigma ...............154	5.1	 What was learned about strategies to reduce weight stigma among health care providers during the development phase? ............................................................................................... 154	5.1.1	 Complex problems require complex, multi-faceted solutions ................................ 155	5.1.1.1	 Challenge attributions about weight as personally controllable ...................... 156	5.1.1.2	 Draw on social consensus to sway opinions .................................................... 157	5.1.1.3	 Awareness raising/self-reflection ..................................................................... 158	5.1.1.4	 Exposure to the challenges of being overweight/obese to evoke empathy ...... 160	5.1.1.5	 Provide a ‘balanced’ anti-medicalization perspective ..................................... 161	5.1.1.6	 Address behaviours and systemic manifestations of stigma ............................ 162	5.1.2	 A ‘balanced’ approach to medicalization ............................................................... 164	5.1.3	 Tailored to emotions – Emotions matter, we need to be prepared for them ........... 167	5.2	 Effectiveness of BalancedView ..................................................................................... 169	5.2.1	 Impact on explicit attitudes ..................................................................................... 170	5.2.1.1	 Changes to specific attitudes ............................................................................ 174	5.2.2	 Impact on implicit attitudes .................................................................................... 178	5.2.3	 Impact on skills and behaviours .............................................................................. 181	5.2.3.1	 Qualitative examples of changes to skills and behaviours ............................... 183	5.2.4	 What is known about when BalancedView was not effective? .............................. 185	5.3	 What was learned about strategies to reduce weight stigma among health care providers during the implementation phase? .......................................................................................... 187	5.3.1.1	 Patient stories and appealing to the heart ......................................................... 189	xi  5.3.1.2	 Appealing to the mind: The influence of evidence .......................................... 192	5.3.1.3	 Exposure to competing ideas on medicalization paired with health-centred education and skill building ............................................................................................ 194	5.3.1.4	 The importance of raising awareness and encouraging self-reflection: But should we use the IAT? ................................................................................................... 198	5.3.1.5	 Importance of a multi-prong approach ............................................................ 202	5.3.1.6	 Pros and cons of a multi-hour, multi-medium, interactive, online course ....... 209	Chapter 6: Discussion and conclusion .....................................................................................214	6.1	 Reflections on research question one ............................................................................. 214	6.2	 Reflections on research question two ............................................................................ 221	6.3	 Contributions ................................................................................................................. 232	6.4	 Methodological reflections ............................................................................................ 235	6.5	 Methodological limitations ............................................................................................ 240	6.6	 Suggestions for future research ...................................................................................... 247	6.7	 Summary and conclusions ............................................................................................. 250	References ...................................................................................................................................252	Appendices ..................................................................................................................................271	Appendix A Scripted reflections ............................................................................................. 271	Appendix B Questionnaires .................................................................................................... 272	B.1	 Pre-course questionnaire ........................................................................................... 272	B.2	 Post-course questionnaire .......................................................................................... 274	B.3	 Follow-up questionnaire ............................................................................................ 279	B.4	 Additional questionnaire questions from pilot .......................................................... 283	xii  Appendix C Interview guides ................................................................................................. 283	C.1	 Interview guide for course participants ..................................................................... 283	C.2	 Group interview guide ............................................................................................... 285	Appendix D Evaluation report ................................................................................................ 287	Appendix E Data tables .......................................................................................................... 314	 xiii  List of tables Table 1 Case study criteria ............................................................................................................ 65	Table 2 Research question one and associated data collection methods ...................................... 78	Table 3 Research question two and associated data collection methods ...................................... 79	Table 4 Gender, age and region of course participants (N = 249) ................................................ 97	Table 5 Profession of course participants (N = 249) .................................................................... 98	Table 6 Interviewee details (n = 10) ............................................................................................. 99	Table 7 Implicit Association Test (IAT) scores .......................................................................... 180	Table 8 Agreement rating: Contributed to skills development ................................................... 182	Table 9 Agreement rating: Helped identify ways to avoid weight stigma in practice ................ 183	Table 10 Impact rating of course components ............................................................................ 189	Table 11 Attitudes about Treating Obese Patients & Perceptions of Weight Bias Among Practitioners: Comparison to Puhl et al. (2014) .......................................................................... 238	Table 12 Contrast of pre- and post-course Fat Phobia scores ..................................................... 314	Table 13 Contrast of pre- and post-course Attitudes about Treating Obese Patients scores ...... 315	 xiv  List of figures Figure 1 Image of methods used in each phase of BalancedView ............................................... 80	Figure 2 Perceptions of weight and health before versus after (N = 249) .................................. 175	Figure 3 Perceptions of focusing on weight loss before versus after (N = 249) ......................... 175	Figure 4 Perceptions of ‘excess fat’ before versus after (N = 249) ............................................ 176	Figure 5 Perceptions of possibility of permanent weight loss before versus after (N = 249) ..... 177	Figure 6 Total available median and mean IAT scores over time .............................................. 181	 xv  List of text boxes Studies using attribution theory with youth .................................................................................. 34	Teachman et al.’s (2003) attributional intervention ...................................................................... 34	Lippa and Sanderson’s (2012) attributional intervention ............................................................. 35	Khan, Tarrant, Weston, Shah and Farrow’s (2017) attributional intervention ............................. 35	Crandall’s (1994) attributional intervention ................................................................................. 36	Diedrichs and Barlow’s (2011) attributional intervention ............................................................ 36	Persky and Eccleston’s (2011) attributional intervention ............................................................. 36	O’Brien, Puhl, Latner, Mir and Hunter’s (2010) attributional intervention ................................. 37	Puhl, Schwartz and Brownell’s (2005) social norms/consensus study ......................................... 38	Zitek and Hebl’s (2007) social norm study ................................................................................... 39	Gumble’s (2012) social consensus study ...................................................................................... 39	Studies using an empathy approach with youth ............................................................................ 40	Rukavina, Li and Rowell’s (2008) empathy intervention with university students ..................... 41	Falker and Sledge’s (2011) empathy intervention with health professionals ............................... 41	Gloor and Puhl’s (2016) empathy and perspective taking study with adults ............................... 42	Gujral, Tea and Sheridan’s (2011) empathy intervention with health professionals .................... 42	Cotugna and Mallick’s (2010) empathy and attribution intervention ........................................... 43	Roberts et al.’s (2011) contact study ............................................................................................. 46	Ciao and Latner’s  (2011) cognitive dissonance study ................................................................. 47	Swift et al.’s (2013) multi-method intervention ........................................................................... 48	Robinson, Bacon and O’Reilly’s (1993) multi-method intervention ............................................ 48	Hague and White’s (2005) multi-method intervention ................................................................. 48	xvi  Gapinski et al.’s (2006) multi-method intervention ...................................................................... 49	Hilbert’s (2016) multi-method intervention .................................................................................. 49	McVey et al.’s (2013) multi-method intervention for health professionals .................................. 51	Frick’s (2007) use of Health At Every Size concepts in a weight stigma intervention ................ 53	Carla and Herman ....................................................................................................................... 163	Tyson and Stephen ...................................................................................................................... 163	Dr. Moffat, Karoline and Mom ................................................................................................... 163	Applying health-centred strategies/case study video observation exercise ................................ 164	 xvii  List of abbreviations AFA   Anti-Fat Attitudes Questionnaire  AFAT   Anti-Fat Attitudes Test  BAOP   Beliefs About Obese Persons Scale BV   The BalancedView course BC   British Columbia  BCMHSUS  BC Mental Health and Substance Use Services  BMI Body Mass Index HAES®  Health At Every Size®*   HCP   Health care provider  IAT   Implicit Association Test  PHSA   The Provincial Health Services Authority  M   Mean (average)  N   Number of participants in the whole study  n   Number of participants in a subset of the study  RQ   Research Question SD   Standard deviation                                                  * Health At Every Size and HAES are registered trademarks of the Association for Size Diversity and Health and used with permission. xviii  Acknowledgements This dissertation was made possible by my wonderful colleagues at the Provincial Health Services Authority who invited me to participate in the BalancedView project. Thank you for the opportunity to contribute to this partnership initiative and for funding the development and evaluation of BalancedView. I am also very appreciative of the funding I received to support my studies from the University of British Columbia, including from the School of Kinesiology, Faculty of Education, UBC Graduate and Postdoctoral Studies and the Centre for Excellence in Indigenous Health.   I am endlessly grateful for the support and encouragement of my supervisor and committee members. To my supervisor, Dr. Patricia Vertinsky, thank you taking me on and for all the time, energy and kindness you have shown me. This dissertation would not have been possible without you. I am so appreciative of the space you allowed me to create a project that was truly transdisciplinary. I am also very beholden to my committee members: Dr. Judith Sixsmith and Dr. Grant Charles. Thank you for providing many hours of your time to both shape and review this dissertation and trouble shoot as necessary. I am also thankful for Dr. Carolyn McEwen for helping to mentor the statistics portion of this project, my colleague and friend Angela Meadows for all the statistics advice and Dr. Marina Niks for the evaluation mentorship. Further, I’d like to thank Dr. Jacqui Gingras, who first introduced me to this field of study. I cannot possibly name all the friends and family who have supported me in this journey. I am grateful for each and every one of you. To my parents – Dennis and Claire – you are my number one supporters and I would not be at this place in my life without you. To my sister Megan, my brother in law Mike and my niece Kaiya, you light up my life. Getting through a PhD was so much more fun with you by my side! To Matthew, thank you for encouraging me to xix  keep going on the PhD and for all the support crossing the finish line. You have made this difficult journey easier with your wonderful presence. Finally, to my fellow PhD students, both current and recently graduated, thank you for your support as we navigated this difficult journey.  xx  Dedication This dissertation is for all the people out there who have ever experienced the devastating impacts of weight stigma. It is also for all the social justice warriors and dedicated academics trying to fix this problem. May we change the world so that coming generations do not face weight bias or discrimination.   1  Chapter 1: Introduction In the last few decades an increasing body of research has shown that weight stigma is a problem across Western societies. Many studies have examined the prevalence and consequences of this stigma, which is seen to manifest in employment settings, educational sectors, the media, interpersonal relationships and health care (Brownell, Puhl, Schwartz & Rudd, 2005; Puhl & Heuer, 2009). For instance, a self-report survey by Puhl, Andreyeva and Brownell (2008) found the prevalence of weight-based discrimination to be comparable to race-, gender- or age-based discrimination. This study focuses upon the topic of weight stigma in health care. Studies have established that health care providers may view overweight or obese patients as weak-willed, self-indulgent, lazy, unmotivated and non-compliant (Bocquier et al., 2005; Brown & Thompson, 2007; Campbell, Engel, Timperio, Cooper & Crawford, 2000; Epstein & Ogden, 2005; Fogelman et al., 2002). Individuals who are the targets of weight stigma often experience consequences to their wellbeing, such as coping through eating (Puhl & Brownell, 2006), lowered physical activity, body image disturbances, disordered eating (McVey, Tweed & Blackmore, 2004; Puhl & Brownell, 2006; Puhl & Heuer, 2009) or avoidance of health care (Amy, Aalborg, Lyons & Keranen, 2006; Olson, Schumaker & Yawn, 1994). More recently, it was found that felt weight stigma contributes to cortisol reactivity (a stress response) (Schvey, Puhl & Brownell, 2014). Among those that study weight stigma and related issues, there is a general consensus that this social and health issue requires significant attention (Cameron & O’Reilly, 2015; Cameron & Russell, 2016; Daníelsdóttir, O’Brien & Ciao, 2010).  There are a number of critical gaps in the literature, however. One overarching issue is that despite many studies exploring the prevalence and consequences of weight stigma, in health care or otherwise, comparably less attention has been paid to defining and conceptualizing 2  weight stigma, as I discuss in my literature review. Among the conceptual studies available, the predominant approach to understanding weight stigma is to view it as a problem of negative attitudes about fatness, which are then presumed to result in unfair treatment of overweight and obese people. A less noted perspective from fat activists and some fat studies scholars is that medicalizing weight and viewing overweight and obesity as diseased may be inaccurate, moralizing and harmful, and a key contributor to weight stigma (Wann, 2009).1 Medicalization is a social process where a condition, behaviour or human variation becomes recognized as diseased (Conrad, 1992; Conrad, 2007). Investigation into the social processes shaping medical understandings of body size and shape is not new (e.g. see Saguy & Riley, 2005; Sobal, 1995; Vertinsky, 2002). However, medicalization as a contributor to weight stigma has only recently been taken up in the weight stigma literature (e.g. Calogero, Tylka & Mensinger, 2016; Cameron & O’Reilly, 2015; McMichael, 2013) and remains an understudied topic. Furthermore, among those who have discussed this issue there is a lack of consensus (see for example Sharma [2012] in contrast to McMichael [2013]). Overall, no one theory has yet adequately explained the existence of weight stigma (UConn Rudd Center, 2017).  Given the lack of understanding around the complexities of weight stigma, it is challenging to formulate effective solutions to the problem. Currently, we know little about how to successfully mitigate weight stigma. Among the studies that have attempted to reduce weight-                                                1 I use the terms overweight and obesity without quotes here, despite the tradition adopted among many fat activists and fat studies scholars to put “overweight” and “obesity” in scare quotes in recognition of the socially constructed nature of these terms and the potential harms resulting from them (Wann, 2009). I opted to go without quotes given the current lack of consensus around the least stigmatizing terms to use (Meadows & Daníelsdóttir, 2016), coupled with the fact that many participants in this study used these terms. However, I believe it remains important to question the social constructions of obesity and, in the words of Meadows and Daníelsdóttir (2016), to “. . . ask ourselves whether the words we use do indeed affirm the respect and human dignity of the target group” (p 3). Also, at times, I use the term fat in this dissertation. I aim to use this word respectfully; with the understanding that fat is neither inherently bad nor good, merely a neutral descriptor that should be reclaimed as such.   3  biased attitudes, many have been unsuccessful in their endeavors (Daníelsdóttir et al., 2010; PHSA, 2013a). Furthermore, few studies have broached the subject of weight stigma reduction in health sectors (PHSA, 2013a). As Lee, Ata and Brannick (2014) point out, “progress has been slow and fraught with doubt as to whether interventions can effectively address a deeply engrained problem” (p 252). This is problematic given the potentially negative implications of weight bias for health and wellness.  Research is thus needed to further elucidate the different ways in which weight stigma can be conceptualized and, more importantly, reduced. Scholarship on how to reduce weight stigma remains particularly important in health care, where little stigma reduction work has yet occurred (Alberga et al., 2016). To address these gaps in the literature I developed two research questions. My first question explored ‘what are the different ways that weight stigma in health care can be conceptualized?’ In order to successfully reduce weight stigma in health care, we need to first have a clear understanding of what it is we are trying to address. In this regard, attention is needed to elucidate the relationship between the medicalization of weight and weight stigma, a topic that is both underexplored and about which there is much disagreement among those who venture to discuss it. My second research question examined ‘what strategies can be employed to reduce weight stigma among health care providers?’ I was interested here in looking at the range of strategies that could be used to mitigate weight stigma in health care and the effectiveness of different strategies in reducing weight stigma. These questions were explored within the context of a participatory case study of weight stigma among health care providers in British Columbia (BC) that was informed by a pragmatic 4  paradigm. In partnership with the Provincial Health Services Authority (PHSA),2 a steering committee of stakeholders with expertise relevant to weight stigma, an advisory committee of health care providers in BC, among others, I helped develop, pilot test, implement and evaluate an online course on weight stigma for health care providers in BC. The focus of the case study was an interactive, five-module course entitled BalancedView (BV) (BC Mental Health and Substance Use Services, 2015).3 The aim of the course, which is still publicly available, is to raise awareness about weight stigma, reduce potential weight-biased attitudes and enhance the competency of health care providers in the province to avoid and reduce weight stigma in their clinical practice. To date the course has been collaboratively developed, pilot tested (n = 8) and launched. At the time my data collection ended,4 249 eligible participants5 had completed the course.  To address my research questions, in keeping with a case study design, I used a number of mixed methods. During the development phase of BalancedView, which was a participatory process with health care providers across the province, I conducted participant observation at meetings concerned with creating BV (an estimated 60 hours of meetings) and kept field notes (58 entries). I also undertook document analysis of 80 project background documents, including: multiple iterations of an evaluation outcomes measurement framework; terms of reference for the committees; scripts for video scenes filmed for the course; 37 meeting minutes; and a scoping                                                 2 Health care in Canada is publicly funded by the federal and provincial government, but provincially delivered. In the Province of British Columbia, the PHSA is a province-wide health authority that works in partnership with the five regional health authorities to promote health, manage chronic disease, prevent and reduce chronic illness and enhance access to evidence-informed health care practice. See www.phsa.ca for more information.  3 See https://balancedviewbc.ca.  4 I stopped collecting data from the BalancedView course in September, 2016, 17 months after the launch.   5 Only individuals who identified as health professionals in BC were included in the study. Students or those outside of BC were excluded.  5  review. In the development phase I also conducted one group interview and co-facilitated a focus group with health care providers involved in the development. In the pilot test stage, I interviewed three participants who took BalancedView and in the implementation stage I interviewed a further 10 course participants. In the implementation phase I also conducted a qualitative document analysis of scripted online reflections completed by participants during the BalancedView course (N = 249) and analyzed pre-course and post-course questionnaires using descriptive statistics (N = 249). Questionnaires were also sent online to a smaller cohort of participants (n = 74) at three- and six-month follow-up. From this group, 56 completed the three-month follow-up and 46 completed the six-month follow-up. The remainder of my dissertation is structured as follows. Chapter two reviews the extant literature related to weight stigma and efforts at its mitigation. I begin with literature on stigma theory and conceptualizations of weight stigma, including related gaps and controversies, particularly in regard to the medicalization of weight. I then consider what is known about weight stigma in health care and its health implications. I conclude with a section on current knowledge on weight stigma reduction and a discussion of those areas where further research is seen to be needed. In chapter three I outline my methodology, including my case study approach, my use of a pragmatic paradigm and a description of the BalancedView case itself. In this chapter I also discuss my methods, sample, analytic techniques and ethical issues. Chapters four and five contain my findings. Chapter four focuses specifically on findings pertaining to research question one. I begin this chapter by addressing how weight stigma was conceptualized by the many stakeholders involved in developing BalancedView and conclude by examining how participants who went on to take the course perceived weight stigma. In chapter five I present my findings related to research question two and discuss ways in which weight stigma among health 6  care providers might be reduced. Chapter five has three main parts. I focus first on the strategies that were considered to reduce weight stigma during the development of BalancedView. Then, I present evidence on the effectiveness of the course in terms of reducing weight bias among course participants. I conclude chapter five by elucidating what elements of the course were perceived as most influential in reducing weight stigma among participants. In my final chapter, chapter six, I provide a broad summary of my findings in relation to the literature, along with my contributions, methodological reflections, limitations, suggestions for future research and concluding remarks. 7  Chapter 2: Literature review  For context, I begin this chapter with an orientation to the broader stigma and prejudice literature. I then discuss weight stigma more specifically, particularly as it pertains to health care and stigma reduction. Studies were identified through key word searches of academic and non-academic, generalized, medical and social science online databases/search engines6, using the terms ‘overweight and obesity’, ‘weight stigma’, ‘weight bias’, ‘obesity and discrimination’, ‘obesity discourse’ and ‘medicalization of weight and obesity’, ‘stigma’ and ‘prejudice’, ‘stigma reduction’ and ‘stigma and labelling’. Searches with additional terms were conducted on an as needed basis to fill gaps in the review.7 I reviewed the reference lists for further manuscripts of relevance when the paper in question was pertinent to my research questions or provided a systematic review.  2.1 How is stigma conceptualized in the research literature?  Goffman (1963), in his iconic book Stigma: Notes on the Management of a Spoiled Identity, conceptualized stigma as “an attribute that is deeply discrediting” (p 13) that reduces a person from whole and normal to tainted and discounted. He identified three main types of stigma: abominations of the body (devalued physical attributes); tribal stigmas (stigma associated with racism); and character blemishes (socially devalued attributes associated with character). He cautioned, however, that attributes in and of themselves are not stigmatizing, but rather it is the meaning ascribed to particular attributes through social interactions and relationships that creates stigma. Despite this, as I discuss later in this review, prior to the new                                                 6 Search engines and databases included PubMed, pubget, the SFU ebrary Library Database, UBC Summons, Google Scholar and Google.  7 For example, ‘attribution theory and weight stigma’, ‘social consensus and weight stigma’, ‘empathy’, ‘compassion’, ‘emotions and weight stigma’ and ‘weight stigma and health care’.  8  millennium, much stigma research did not focus on the social processes or factors that enable stigma. This is an issue that is somewhat mirrored in the weight stigma literature.  While Goffman (1963), a sociologist, is often credited with bringing stigma into the purview of social science, Allport, a social psychologist, is viewed as similarly influential in the study of prejudice. Allport (1954) contended that it is a natural part of human life and social order to develop categories that help us cognitively process the huge amount of information and stimuli in our everyday lives. This process of categorization means that we ‘pre-judge’ or stereotype people according to which category we cognitively identify them with. Although some degree of pre-judgment is normal, he argued, overly generalized pre-judgments, when coupled with antipathy towards a group, lead to unfair prejudice (Dovidio, Glick & Rudman, 2005). Although initially studied as a separate concept from stigma, today prejudice and stigma research have some overlap (as discussed below) and, thus, insights from the prejudice field are important to understand.  In the decades since Goffman (1963) and Allport (1954), two related yet somewhat separate bodies of literature have developed around the issues of stigma and prejudice. Although Goffman (1963) discussed tribal stigma, research on race has been mostly taken up in the prejudice literature, alongside ideas of group domination and exploitation. Stigma research, by contrast, has focused more on the idea of deviance from norms, particularly as it pertains to deviant behaviour (Bos, Pryor, Reeder & Stutterheim, 2013; Phelan, Link & Dovido, 2008) and on stigmatizing attitudes about socially devalued attributes (Cameron & O’Reilly, 2015; Link & Phelan, 2001). Perspectives are divided concerning the conceptual differences between these two terms, if any. Some authors appear unconcerned by potential conceptual discrepancies, using the terms prejudice and stigma interchangeably in the same texts. Bos et al. (2013) make the 9  distinction that stigma research tends to be conceptually centered on the idea of deviance and reactions to deviance, whereas prejudice is essentially about pre-judgment, irrespective of whether that group is deviant from normative societal expectations. Phelan et al. (2008), in a review of the theoretical literature on stigma and prejudice, argue that prejudice can be best seen as the attitudinal component of a broader stigma process that involves power, labelling, prejudicial attitudes, stereotypes and discrimination. While some prejudice scholars seem to agree with this notion of prejudice encapsulating attitudes specifically, others consider prejudice to refer to attitudes and behaviour (Daníelsdóttir et al., 2010). Stigma scholars similarly sometimes focus exclusively on attitudes, while at other times on both attitudes and behavioural manifestations of attitudes, such as discrimination. What is important, regardless of the approach taken, is that there is clarity of concept (Brownell et al., 2005). In looking at how stigma has been conceptualized since Goffman (1963), one can detect a notable shift in the literature around the new millennium (Cameron & O’Reilly, 2015). Prior to 2000, a major emphasis in stigma research was on micro-level interactions and on stigmatizing attitudes held by individuals towards others with particular deviant attributes. This work largely focused on cognition and affect and was primarily undertaken by social psychologists (Link & Phelan, 2001). Since 2000, there has been an increased focus on macro-level factors that promote stigma (Yang et al., 2007). In particular, there has been an emphasis on the significance of labels and power to the stigma process. Regarding labels, Link and Phelan (2001, 2013) contended that it is through the social process of labelling that other aspects of stigma such as stereotypical attitudes and discrimination occur. Link and Phelan (2001) and Link, Yang, Phelan and Collins (2004) highlighted that the concept of labels is different from that of attributes, in that labels allow us to consider the social construction of characteristics seen as different or deviant. As 10  they aptly pointed out, attributes that are considered stigmatizing are not necessarily naturally stigmatizing conditions, as can be seen by the huge variance in what is considered a stigmatizing attribute by time and place, but rather socially constructed as such. Regarding power, Link and Phelan (2001) further discussed that both labelling and stigma are linked to ideas of power. As they showed, it takes power to label groups of people as fundamentally different and lesser than others.  Although the study of stigma and prejudice has developed considerably since the seminal works of Allport (1954) and Goffman (1963), theoretical scholarship on weight stigma or prejudice is in its infancy in comparison (Daníelsdóttir et al., 2010). Inquiry into weight stigma began in earnest in the 1990s (Brownell, Puhl, Schwartz & Rudd, 2005), although this was predated by second wave feminist scholarship on ‘the tyranny of slenderness’ and late 1960s and 1970s fat activist work (McMichael, 2013). Today, there is a wide-ranging literature within psychology (Puhl & Heuer, 2009), sociology (Saguy, 2013), fat studies (Murray, 2008), feminist scholarship (Fikkan & Rothblum, 2012), health fields (McVey at al., 2013), rhetorical studies (McMichael, 2013), history of health and physical education (Vertinsky, 2008), critical weight studies (Rich, Monaghan & Aphramor, 2010) and other areas pertinent to weight stigma. While much has been learned about the prevalence and consequences of weight stigma in the last few decades, there are fewer studies focused on generating theory and conceptual clarity about weight stigma specifically. Additionally, the concept itself is theorized in diverse – and sometimes conflicting – ways, as discussed in the next section. 2.2 How is weight stigma conceptualized in the research literature?  What is weight stigma? How is the concept defined? While there is extensive work on the prevalence and consequences of the problem, there has been comparably less attention to 11  elucidating what is meant by ‘weight stigma’. In some texts the nature of the problem is assumed, with few definitions provided  (e.g. Puhl, Moss-Racusin & Schwartz, 2007; Puhl, Peterson & Luedicke, 2013). Overall, the field of weight stigma would benefit from attention to problem definition. As Lee et al. (2014) point out with respect to the field of weight stigma reduction interventions, “A clearer definition of weight bias and establishment of cut off points for clinically significant levels of weight bias . . . would benefit this area of study” (p 258). An article by Alberga, Russell-Mayhew, von Ranson, McLaren, Ramos Salas and Sharma (2016), based on discussions at the 2015 Canadian Weight Bias Summit, highlighted that one priority for ongoing weight stigma research is “. . . to identify the root causes and definitions of weight bias through qualitative and quantitative research” (p 1208). The importance of this conceptual work is particularly integral to stigma reduction efforts, as problem definition is central to problem solving (D’Zurilla, Nezu & Maydeu-Olivares, 2004). At present, among extant studies dedicated to conceptualizing the problem, the predominant trend is to utilize a psychosocial, primarily attitudinal approach, much like stigma research prior to the new millennium. Within this, attribution theory and social consensus theory are common explanations for weight stigma.  2.2.1 Psychosocial approaches to weight stigma Social psychologists have a long history of studying attitudes, which can be conceptualized as assumptions and reactions, either favourable or unfavourable, that predispose individuals to a particular course of action or behaviour (Fishbein & Ajzen, 1975). Building on this tradition of attitudinal research, scholars working within social psychology often refer to weight stigma as an issue of negative attitudes (or beliefs) about the attribute of obesity (e.g. viewing fat people as lazy, lacking in willpower), which may lead to weight-based discrimination (Puhl & Heuer, 2009). While some weight stigma studies distinguish between 12  attitudes and beliefs as independent constructs, many do not (Lee et al., 2014). Attitudes can also be conceptualized in two ways, as implicit or explicit. Implicit attitudes are automatic reactions that are emotionally based and do not require conscious thought, thus may be inconsistent with one’s belief system. Explicit attitudes are our assumptions and reactions that are consistent with our belief system and require consideration of information and critical thinking, thus are endorsed as ‘true’ (Watts & Cranney, 2009). Reportedly, implicit weight-biased attitudes are more resistant to change than their explicit counterparts (Teachman et al., 2003). Further, explicit and implicit bias, while related, have been found to be only moderately correlated (Puhl, Phelan, Nadglowski & Kyle, 2016). As is shown by this distinction between implicit and explicit attitudes, the explicit dimension pulls on the construct of beliefs, perhaps explaining why some scholars do not distinguish between attitudes and beliefs. As will be evident later, the approach taken in BalancedView and my study is to conceptualize attitudes and beliefs as similar, with beliefs reflecting the more explicit dimension of attitudes. Next, specific theories of weight stigma aligned with the attitudinal literature are discussed.  Attribution theory (Crandall, 1994) is a social psychology theory extensively studied in the context of stigma. Proponents of this theory argue that in cultures where individualism is highly valued, stigma results when individuals attribute differences in outcomes to individual choice. The argument is that overweight and obesity are thus stigmatized when individuals attribute higher weights to poor personal choices (Puhl & Brownell, 2003). Social norm approaches to stigma, again based in social psychology, emphasize that stigma may be created or mitigated through social norms which influence attitudes. Social consensus theory builds on social norm theory insofar as it suggests that social norms are particularly likely to be affected 13  and influenced by ‘in group’ members: those who hold socially influential and respected positions in society (Puhl, Schwartz & Brownell, 2005).  In addition to these psychosocial theories, other attitudinal explanations of weight stigma have also been proposed. For instance, O’Brien, Daníelsdóttir, Ólafsson, Hansdóttir, Fridjónsdóttir and Jónsdóttir (2013) looked at the relationship between appearance concerns, disgust and anti-fat prejudice (measured by the Anti-Fat Attitudes scale). They found that physical appearance worries and disgust were related to anti-fat attitudes for women. It was unclear whether disgust of fatness was a result of pathogen avoidance (an evolutionary explanation for stigma associated with fear of disease) or to do with sociocultural perceptions of fatness as a moral breach. Others have also drawn on the disgust literature, looking at the relationship between disgust sensitivity (individual propensity to react to elicitors of disgust) and weight-biased attitudes, as well as fear of perceived pathogen threats and how this contributes to stigmatizing attitudes (Lieberman, Tybur & Latner, 2011).  As another example of weight stigma theory in the attitudinal tradition, Azétsop and Joy (2011) suggest that weight-biased attitudes result in contexts where individualistic values are paired with a general cultural tendency towards anti-intellectualism (e.g. prioritizing common sense over critical thinking or science). Others have approached weight stigmatizing attitudes via an exploration of determinants of these attitudes in quantitative studies, with potential determinants including Body Mass Index (BMI) and education levels (Davison & Birch, 2004; Hilbert, Rief & Braehler, 2008). A limited amount of studies have looked at implications of the labels ‘overweight’ and ‘obese’ and their relationship to stigma. Vartanian (2010) found that the language of ‘obese people’ evoked stronger negative reactions than the language of ‘fat people’ among a sample of undergraduates (N = 425), suggesting that weight stigma may be promoted 14  through labelling, something that I take up in this study. Hunger and Tomiyama (2014) also found that the label of ‘too fat’ increased the likelihood of being obese by BMI standards close to 10 years later, pointing to a possible connection between labelling and weight gain.  The most prominent example of an attitudinal approach to weight stigma comes from the Rudd Center for Food Policy and Obesity, a research institute formerly located at Yale University that transitioned to the University of Connecticut in 2015. The Rudd Center has a dual mission – they strive to both prevent obesity and reduce weight stigma  (Rudd Center, 2013a) – and have produced a huge volume of literature related to these topics.8 Rudd Center colleagues claim that we do not know enough about why weight stigma exists (Puhl & Brownell, 2003; UConn Rudd Center, 2017) and they turn primarily to attitudinal psychological theories to explain weight stigma, including attribution theory and social consensus theory (described above) (Puhl & Heuer, 2009). A study by Puhl, Latner, O’Brien, Luedicke, Daníelsdóttir and Forhan (2015) found that across Canada, the United States, Iceland and Australia, attributions of behavioural causes of overweight and obesity predicted higher levels of attitudinal weight bias. Additionally, the Rudd Center has considered how media portrayals of obesity can impact attitudes (McClure, Puhl & Heuer, 2011). The Rudd Center, however, does not limit their work to an attitudinal focus. Rather, they often emphasize the inequitable behaviours (e.g. bullying) and outcomes (e.g. lack of employment opportunities) likely to result from weight-biased attitudes.                                                  8 Their website had over 35 publications on the subject between 2007 and 2014 alone (Rudd Center, 2013b). They have written several systematic reviews and conducted experiments and surveys to establish the prevalence of weight stigma and its consequences. Additionally, relative to other scholarship on weight stigma, the Rudd Center is highly visible in the media (Rudd Center 2013c). 15  Interestingly, while attitudinal frameworks of weight stigma dominate – and there seems to be little disagreement in the literature that we have a problem of weight stigmatizing attitudes in contemporary Western society – it is not always clearly specified what, exactly, is considered to be a weight-biased attitude. One issue is that in the weight stigma literature the construct of attitudes is sometimes taken for granted and not defined. Another issue is that what attitudes are considered to be weight-biased vary from study to study and between attitudinal measures, though there are often commonalities, particularly around willpower (e.g. a commonly cited weight-biased attitude is to view fat people as lazy or lacking in willpower). It is also unclear what the relationship is between weight-biased attitudes and other aspects of weight stigma, such as discriminatory behaviours. The premise is that attitudes may influence behaviour, however the relationship between weight-biased attitudes and behaviours is still underexplored (O’Brien, Latner, Ebneter & Hunter, 2013).  Attitudinal work on weight stigma, while immensely valuable in terms of helping us understand the prevalence of weight-biased attitudes, is often aligned with notions of overweight and obesity as health problems. This is exemplified by the Rudd Center’s dual mission to prevent and reduce obesity and weight stigma. This is in direct contrast to the contentions of fat activists who believe that the construction of overweight and obesity as health problems is not only inaccurate but moralizing and harmful (Wann, 2009). Below, I discuss other ways that weight stigma can be understood beyond the dominant psychosocial framework, including the claims of fat activists.  2.2.2 Other approaches to weight stigma  Outside of the dominant psychosocial conceptual work on weight stigma, there have been far fewer studies focused specifically on conceptualizing ‘weight stigma’. However, there is an 16  extensive scholarship on related topics, such as critical theory on the social construction of the obesity epidemic (e.g. Boero, 2012; Gard & Wright, 2005; Kwan, 2009) and critical weight studies (Rich, Monaghan & Aphramor, 2010). We have also seen long standing activist efforts focused on ending discrimination against fat people, for example through the decades old National Association to Advance Fat Acceptance (NAAFA) (NAAFA Inc., 2016). In recent years we have also seen the emergence of the new field of fat studies (Rothblum & Solovay, 2009). As Wann (2009) argues in the foreword to the Fat Studies Reader, fat studies can be defined “…in part by what it is not. For example, if you believe that fat people could (and should) lose weight, then you are not doing fat studies” (p ix). While fat studies allows for a diverse and interdisciplinary study of fatness and weight extending well beyond weight stigma, some scholars in this field like Farrell (2011) and McMichael (2013) have also focused specifically on weight stigma.   Farrell’s (2011) book, Fat Shame: Stigma and the Fat Body in American Culture, explores the roots of negative contemporary American ideas about fatness. Using Goffman’s (1963) three-fold typology of stigma, Farrell suggests that fatness is at once an abomination of the body, a character stigma and a tribal stigma used to mark racialized bodies as uncivilized. While she draws briefly on Goffman, her primary analysis centres around the cultural studies concept of ‘citizenship’ as it relates to weight stigma. Her research uses a historical analytical perspective and explores cultural artifacts (e.g. magazines, journals, post cards and propaganda) that have been used to represent the fat body from the mid 1800s onwards. Her thesis is that current stigmatizing American concepts about fat are rooted in historical ideas about race, gender, civilization and citizenship. She argues, for example, that in a pre-industrialization age, with a highly class stratified society, food was scarce and only the elite upper class could afford 17  to consistently eat well. Thus, fatness was by and large associated with affluence, wealth and health and was idealized to some extent. With industrialization, food and resource availability increased and reportedly more people in the lower classes were able to gain weight. With industrialization came some upward mobility, the emergence of a growing middle class and the development of racial and immigrant tensions with the influx of immigrants from Eastern Europe. The upper class, in an effort to distance themselves from this upwardly mobile, not-as-purely-white middle class, began to look down on fatness. They made the age-old connection between fatness and greed and associated middle and lower class plumpness with an inability to handle the temptations of modern life. In an increasingly capitalist society, fatness began to be framed as indicative of an inability to manage desires of excess and was labelled as uncivilized, with fat people thus constructed as less deserving of citizenship. This was reportedly justified from a scientific perspective through the popularity at the time of social Darwinism and the increasing use of technologies to measure bodies, such as weight scales. Hierarchies of civilization were created and those who were the leanest and thinnest were constructed as the most evolved and ‘fittest’.  Farrell’s (2011) book reminds us that constructs about fatness are intimately related to how we think about poor people, women and racialized groups. While she also provides a compelling look at historical and ‘big picture’ factors that have encouraged weight stigma, she does not look in-depth at the ways in which medical ideas about obesity have contributed to weight stigma. The relationship between the medicalization of weight and stigma is an important area of exploration given the articulations by fat activists and fat acceptance community members (McMichael, 2013; Wann, 2009) that framing overweight and obesity as a health crisis leads to the demonization of fat people.  18  McMichael (2013), a self-identified fat activist, fat acceptance community member and fat studies scholar takes this issue up at length. She builds on the contentions of fat activists who have long been arguing that seeing fat as a medical problem is moralizing and harmful. To quote an influential fat activist, Wann (2009) notes that:  “[o]verweight” is inherently anti-fat. It implies an extreme goal: instead of a bell curve distribution of human weights, it calls for a lone, towering, unlikely bar graph with everyone occupying the same (thin) weights. If a word like “overweight” is acceptable and even preferable, then weight stigma becomes accepted and preferred. (Wann, 2009, p xii) While it is common within fat activist circles to see conversations about medicalization contributing to weight stigma (e.g. within the ‘Fatosphere’9) and in a small, but growing number of articles (e.g. Calogero, Tylka & Mensinger, 2016; Goldberg, 2014), McMichael’s (2013) work is one of the few in-depth pieces that I am aware of that focuses on connecting theories of weight stigma and prejudice to ideas about fat as a medical problem.10 In her book, she develops a theory of weight-based oppression based on bell hooks’ ideology of oppression.  Her argument is that fat prejudice can be understood through the lens of rhetoric and hooks’ notion of domination (hooks, 1989 as cited in McMichael, 2013). McMichael defines rhetoric as any persuasive act of communication: written, verbal, visual or otherwise. She                                                 9 A linked community of fat acceptance and activist bloggers. 10 Of note, many other scholars not focused specifically on stigma have examined the history of the medicalization of weight. Such exploration of the social processes shaping medical understandings of body size and shape is not new (see Saguy & Riley, 2005; Sobal, 1995; Vertinsky, 2002). For instance, Vertinsky (2008) discusses how multiple social processes informed medical understandings of weight, including the insurance industry’s long use of weight scales and height and weight tables to predict population health and illness.  19  believes that rhetoric is the essential medium through which people both develop and resist prejudice. Hence, she conceptualizes prejudice as a pre-judgment or opinion that is not based on reason or actual experience, much like Allport’s (1954) historic conceptualization. She sees these pre-judgments as central to stigma, discrimination and oppression for fat people. To better understand how prejudice develops she relies on hooks’ contention that patriarchal and other oppressive forces create systems of oppression and domination, in which the oppressed internalize the values of the dominant. Connecting this to rhetoric, she sees rhetorical language and communication as the vehicle through which dominant values are developed, internalized and reinforced.  McMichael further contends that fat prejudice is perpetuated in Western society due to two persuasive myths: that fat is unhealthy and that anyone can be permanently thin if they eat well and exercise right. She suggests that equating thinness with health and fatness with disease is inherently prejudicial in that by assuming we can judge health status (and health behaviours) simply by someone’s size is a pre-judgment. Due to widespread dissemination of these two messages about fat as unhealthy and changeable, via the media, interpersonal relationships and in health care, fat people can be persuaded to internalize the idea that their bodies are sick and that they are that way because they lack willpower. This internalization is reinforced, she articulates, through the historical and continuing influence of religious and moral perceptions of gluttony as sin, and the cultural value we place on restraint and willpower. Thus, fat people come to believe that their bodies and morals are in some way lacking and inferior, and learn to police their bodies in various ways. When and if their bodies stay fat, despite self-policing, fat people further buy into the idea that they are inferior and deserve to be treated badly. 20  In my study I explored McMichael’s (2013) contention that medical beliefs about overweight and obesity lead to fat prejudice. However, rather than utilizing hooks’ ideology of oppression, I grounded the idea that medical beliefs may lead to prejudice more within a model of stigma than oppression (or prejudice). ‘Weight stigma’ is a topic of great social salience at the moment. It is now recognized as a public health issue and governments and academic bodies alike are taking interest. Therefore, it is possible that working within a stigma framework is more likely to have social impact.  In addition to Farrell’s (2011) and McMichael’s (2013) work specific to weight stigma, there is other critical theoretical scholarship focused, not on conceptualizing weight stigma per se, but on related constructs, including: Gard and Wright (2005); Boero (2012); and Lupton (2013). Saguy’s (2013) work is also particularly relevant as, while her main aim is not to conceptualize weight stigma, she connects her work to attitudinal constructs of weight stigma. Each is discussed below.  Gard and Wright (2005) show that ‘obesity epidemic’ thinking is as much about moral anxieties as it is about science. They illustrate how anxieties about some of the ‘scourges of modern life’ – such as an increased reliance on technology and the ‘couch potato’ phenomenon – are connected to ideas about fatness (despite a lack of scientific evidence to support this) and that these anxieties lead to an enhanced sense of panic about obesity. They demonstrate that the science underpinning the ‘obesity epidemic’ is contradictory and inconsistent, and that there is a considerable amount of uncertainty about obesity as a legitimate public health threat.  Boero (2007) similarly strives to understand the process through which an ‘obesity epidemic’ was socially constructed. She argues that the public health emphasis on the obesity epidemic is as much about morals and a moral panic as it is about science. She sees moral 21  entrepreneurs – individuals or groups who advance a particular moral claim – as partially responsible for obesity being designated as a disease and a threat in American public health. Lupton (2013) explains this sense of panic well. She suggests that in a culture where control and containment are highly valued, such as ours, fat bodies are seen as challenging these values and transgressing the culturally normative shape that bodies (especially female bodies) are supposed to take. As with Gard and Wright (2005), she perceives this panic as being largely fuelled through an increased reliance on BMI, as well as through the large volume of experts and news sources reporting on the health crisis of obesity.  Saguy (2013), a sociologist in the social constructionist tradition and an influential fat studies scholar, argues that medical frames about obesity and weight are central to the negative social experiences of fat people in contemporary Western society. Saguy’s (2013) analysis, while not centred entirely on weight stigma, aims to elucidate the various frames surrounding fatness, explore how certain frames come to dominate over others and investigate the social implications of these frames, one implication of which is stigma.  Using the notion of framing, Saguy argues that there are two main types of frames: problem frames and blame frames. Problem frames are frames that define something as a social problem (or conversely, as socially unproblematic). Blame frames are used once something is defined as a social problem to explain why the issue exists. She articulates three main trends in problem framing around weight: 1) an immorality frame, in which fatness is constructed as a moral problem; 2) a medical frame, in which fatness is seen as an individual health problem; and 3) a public health crisis frame, in which fatness is framed as a population health problem with economic implications. Through a textual analysis of news reporting on obesity, she locates the power of the three problem frames with historic roots in religious, medical, public health and 22  scientific authority. In contemporary society, groups promoting problem frames on weight have a higher degree of social, symbolic, economic and bodily capital than those in favour of non-problem frames (e.g. fat acceptance activists). She also points to the oversimplification of weight science in which some findings are emphasized over others by researchers and then uncritically promoted and further dramatized by the news media who accept such science as objective.  Regarding blame frames, Saguy similarly locates three main frames: 1) a personal responsibility lens; 2) a biological perspective; and 3) a sociocultural frame. Each of these blame frames, she contends, take for granted that there is an obesity crisis. Thus, even though an environmental, sociocultural frame should be less stigmatizing than a personal responsibility explanation, by virtue of taking the problem of obesity as a given, all causal discussions of obesity reinforce ideas about fat as problematic. To support the argument that there are negative implications of medical frames, she presents findings from seven controlled experimental studies. Each experiment presented participants with actual news articles on weight, after which they completed attitudinal measures. Control groups were exposed to articles with non-problem frames (e.g. content that suggested that people can be healthy at a range of weights if engaging in healthy eating and activity). The other groups were shown articles with various problem and blame frames. She found that participants exposed to medical or public health problem frames were more likely to be weight-biased than those exposed to non-problem frames. In my study, I build on Saguy’s research investigating the relationship between medical frames and weight stigmatizing attitudes. Through my first research question,11 I consider how Saguy’s (2013) work on framing relates to Link and Phelan’s (2001) ideas about labelling and                                                 11 Research question one: ‘What are the different ways weight stigma in health care can be conceptualized?’ 23  stigma. I believe this is important given that fat studies scholars, fat activists and proponents of a health-centred approach can often be heard arguing for the central role of the pharmaceutical industry in constructing ‘overweight’ and ‘obesity’ as health problems and evidence on the harms of this construction.  Fat studies scholars Bacon (2010) and Saguy (2013), for example, claim that one reason for why we think higher weights are a health problem relates to the influence of the pharmaceutical industry. A key example of this occurred in the late 1990s, when the World Health Organization adjusted the Body Mass Index (BMI) cutoffs for overweight. As Saguy, Gruys and Gong (2010) allege, the International Obesity Task Force (IOTF) “. . . paid for a staffer to draft the World Health Organization (WHO) report that advocated lowering the [overweight BMI] cut-off [from 28 to 25] ” (p 591), a move which caused  “. . . 29 million Americans to become ‘overweight’ overnight” (p 591). Notably, the IOTF was an anti-obesity lobbying group funded by the manufacturers of various weight loss drugs. This suggests that the medicalized nature of overweight and obesity is partially related to the influence of the pharmaceutical industry and fuelled by economic interests.  Exploring how medicalization may contribute to weight stigma is particularly important given scholarship that calls into question trustworthiness and accuracy of research finding a relationship between weight and health and the potential harms of weight normative approaches (Tylka et al., 2014). Bacon and Aphramor (2011), for example, interrogated the relationship between weight and health and examined the implications of focusing on weight as a measure of health. They argued that the relationship between weight and health is not as simple as the dominant weight-focused paradigm suggests. Specifically they contended that people in the ‘overweight’ BMI category have longer life expectancies than those in the ‘normal’ category. 24  They similarly put forward data on the ‘obesity paradox’: in some cases obesity appears to be protective and promote survival in the face of problems like cardiovascular disease or type two diabetes. Thus, they suggest that weight-focused health care mistakenly targets overweight and obese people as requiring intervention. They similarly contend that focusing on weight loss as a means of improving health undermines rather than improves health. They provide evidence illustrating the unlikelihood of long-term weight loss. Most people who lose weight regain the weight and this process of ‘weight cycling’, they contend, is more harmful to health than maintaining a weight that is higher, but stable. Focusing on losing weight also increases the likelihood of disordered eating. Bacon and Aphramor (2011) thus suggest that what is required in health care is an emphasis on diet and fitness rather than on weight, as studies have shown these factors to be important to health outcomes, regardless of weight status (Bacon et al., 2002; Bacon, Stern, Van Loan & Keim, 2005).  Building on Bacon and Aphramor (2011), in my master’s work (O’Reilly & Sixsmith, 2012), I took this argument further and suggested that promoting weight-centred approaches to health within health systems should be considered systemic discrimination. I argued that it is discriminatory to promote weight loss within health sectors given the evidence showing that this focus leads to weight cycling and associated harms and eating disorders. However, while I argued that the medicalization of weight was discriminatory, I provided little clarity on what I meant by stigma and discrimination or how and why medicalization and the weight stigma construct were related.  While I am interested in taking the claims of fat activists seriously and advancing a debate about how medicalization relates to weight stigma, the idea of medicalization as weight stigmatizing is a radical concept to many and remains on the margins of weight stigma discourse. 25  The Rudd Center and other dominant attitudinal research on weight stigma, for example, do not address this topic. Most weight stigma research actually medicalizes weight, insofar as arguing for the simultaneous need to reduce population weights and reduce weight stigma (Calogero, Tylka & Mensinger, 2016). Some obesity scholars also seemingly disagree with the fat activist perspective. Sharma (2012), a Canadian obesity scholar and medical doctor has argued – in direct contrast to the premise of fat activists, like Wann (2009) and McMichael (2012) – that the medicalization of obesity is necessary to avoid weight stigma. He perceives that to view obesity12 as a disease may help shift blame from individuals and provide training for health care providers to assess who with excess fat is diseased versus healthy (Sharma, 2012). A similar perspective is promoted by Blackburn (2011), who acknowledges some shortcomings of medicalizing weight, yet nonetheless advocates for medicalizing obesity as a means of reducing weight discrimination in health sectors. To date, however, such controversy remains sidelined within weight stigma research. Clearly there are competing perspectives on whether and how medical frameworks of weight relate to weight stigma. However, such debate is rarely aired. The lack of attention and debate around this issue is likely because we live in an era where ‘obesity epidemic’ thinking has infiltrated most aspects of our culture, including our medical and academic institutions. This means that the prevalence and implications of weight stigma are not fully understood. To date, what we know about the prevalence and implications of weight stigma primarily relates to concepts of weight stigma as an attitudinal problem with consequences for individuals who are stigmatized. Attitudes or beliefs about weight as a medical problem are not                                                 12 Sharma (2012) defines obesity as a problem where excess fat impacts or threatens health. 26  often discussed as part of this. In the next section I discuss what is currently known about the prevalence and consequences of weight stigma, particularly in health care.  2.3 Prevalence and implications of weight stigma  In contrast to the smaller amount of conceptual work on weight stigma, a huge volume of literature looks at the prevalence or consequences of the problem. Most of this work uses attitudinal measures of weight bias and self-report or experimental measures for discrimination. As Puhl and Heuer (2009) demonstrate in a systematic review of weight stigma literature, numerous studies have found that weight stigma is prevalent across society: in employment and educational settings, the media, interpersonal relationships and in health sectors. For instance, one self-report study with 2,449 overweight and obese women found that 54 percent of participants reported experiencing weight stigma from their colleagues and 43 percent reported weight stigma from their managers/employers (Puhl & Brownell, 2006). The authors also found that 32 percent of overweight and obese adult women reported being subject to weight stigma by a professor or teacher. They further found that interpersonal relationships and, more specifically, family members, friends or spouses were significant sources of weight bias. For example, 72 percent of participants reported having experienced weight stigma from family members (Puhl and Brownell, 2006).  Weight stigma may also result in discrimination against heavier people. In a self-report survey conducted by Puhl, Andreyeva and Brownell (2008) it was found that the prevalence of weight-based discrimination is now comparable with race-, gender- or age-based discrimination. In this study, data were analyzed from a survey of 2,290 American adults looking at sources of perceived discrimination, including: age, gender, race, ethnicity or nationality, religion, physical disability, height or weight, sexual orientation or other. The authors found that among women, 27  height- or weight-based discrimination was the third most prevalent form of discrimination, preceded only by discrimination on the basis of gender and age. Among men and women, height or weight discrimination was the fourth most prevalent form of discrimination, with discrimination based on age, gender and race occurring more often. In this study, among those with a Body Mass Index (BMI) of 35 of higher, 40 percent reported weight-based discrimination. Consequences of felt weight stigma that are often cited include emotional eating, lowered physical activity (Puhl, Moss-Racusin & Schwartz, 2007) and body image disturbances (Friedman et al., 2005). Additionally, positive correlations have recently been found between exposure to weight stigma and greater cortisol reactivity – a biological indicator of chronic stress (Schvey, Puhl, & Brownell, 2014).  2.3.1 Weight stigma in the health sector  Along with the exploration of weight stigma in relation to employment, education, the media and interpersonal relationships, the issue has also often been investigated among health professionals (Fruh, Nadglowski, Hall, Davis, Crook & Zlomke, 2016; Puhl, Phelan, Nadglowski & Kyle, 2016). Numerous studies, including qualitative studies (Brown & Thompson, 2007; Epstein & Ogden, 2005; Kirk et al., 2014), quantitative self-report studies (e.g. Campbell, Engel, Timperio, Cooper & Crawford, 2000; Fogelman et al., 2002) and experimental studies (e.g. Hebl & Xu, 2001), have explored health care providers’ (e.g. nurses or doctors) perceptions of fat patients and suggested that they hold stigmatizing attitudes towards obese patients, viewing them as weak-willed, lacking in discipline or lazy. As one questionnaire-based study of 620 primary health care doctors found, over 50 percent of physicians perceived obese patients as unattractive, ugly and non-compliant and another third viewed them as weak-willed and lazy (Foster et al., 2003). A semi-structured, interview-based study with 15 nurses found that although participants 28  attempted to avoid weight-based stereotypes and recognize the complexity of obesity, they nonetheless associated obesity with non-compliance by patients (Brown & Thompson, 2007). In another interview study with 21 primary health care physicians, doctors perceived obesity to be a personal responsibility of the patient and viewed obese clients as non-compliant and unwilling to engage in healthier lifestyles (Epstein & Ogden, 2005). Sabin, Marini and Nosek (2012) found that among a sample of 2,284 physicians, weight bias was as problematic and prevalent as among the broader public.  Overall, research has suggested that many doctors may view overweight and obesity as the result of laziness, not enough physical activity and too much food (Bocquier et al., 2005; Foster et al., 2003). They may also consider obese patients to be weak-willed, sloppy, unattractive (Foster et al., 2003), self-indulgent (Bocquier et al., 2005) and non-compliant (Thuan & Avignon 2005). Such attitudes have also been found among nurses (Pervez & Ramonaledi, 2017). For instance, one study found 69 percent of nurses believed that food and exercise decisions cause obesity and one third of nurses attributed this to poor willpower (Brown, Stride, Psarou, Brewins & Thompson, 2007). A recent study by Kuehl, Kirk, Dumas and Kyle (2017) using online surveys found that only 28.7% of health care professionals in Canada (n = 576) and 29.2% (n = 676) in the US perceived obesity as a problem of bad personal choices. These results suggest that weight-biased attitudes about fatness as a personal choice may be shifting. However, other recent research comparing levels of weight bias among health care and obesity professionals found comparable levels of bias between 2001 and 2013, although implicit weight bias appeared slightly lower, while explicit weight bias seemed higher in the more recent sample (Tomiyama et al., 2015). Recent research also shows that weight-biased attitudes may also exist among some eating 29  disorder practitioners (Puhl, Latner, King & Luedicke, 2014), a group often assumed to have enhanced sensitivity to issues around weight. While the primary purpose of my study is not to analyze the scope of weight-biased attitudes among health care providers, my study does add to what is known about this topic in BC, Canada through the use of questionnaires to measure weight-biased attitudes among health care providers enrolled in BalancedView, the course of focus in my case study.  Unfortunately, weight stigmatizing attitudes among health care practitioners may have implications for how fat people are treated in and able to access the health care system. One study found that 31 percent of nurses would prefer not to care for obese patients (Maroney & Golub, 1992).13 Another study found that eating disorders are underdiagnosed and undertreated among obese youth, as practitioners are less likely to recognize heavier youth as having disordered eating (Sim, Lebow & Billings, 2013). Because of perceived actual or potential weight stigma in health arenas, fat people may avoid or delay seeking medical attention for fear of how they will be treated (Drury & Louis, 2002; Olson et al., 1994; Puhl & Heuer, 2009). For instance, Amy et al. (2006) found that obese women who experienced disrespectful treatment from doctors due to their weight or received unsolicited weight loss advice during health care encounters were more likely to delay or avoid cancer screening tests. Perceived weight bias in health care may also result in patients being less likely to trust their primary health care providers (Gudzune, Bennett, Cooper & Bleich, 2014). An interview study (N = 24) with women in                                                 13 Preferences among health care providers to not provide care for heavier patients may be a direct result of weight stigmatizing attitudes or may stem from environmental stigma, such as a lack of medical equipment that can accommodate people of size and enable health care providers to safely and easily care for larger people.  30  reproductive health care found that health care providers sometimes refuse to provide care to patients on the basis of weight (Bombak, McPhail & Ward, 2016).  Other consequences for individuals who perceive weight stigma, as discussed, include emotional eating, lowered physical activity (Puhl, Moss-Racusin & Schwartz, 2007), body image disturbances (Friedman et al., 2005) and greater cortisol reactivity (Schvey, Puhl, & Brownell, 2014). Feelings of shame may also arise as a result of felt stigma in health care encounters. Kirk et al. (2014) used qualitative interviews to explore experiences of individuals that identified as overweight and found that the dominant advice often present in health care to be more active and eat less was associated with feelings of blame and shame among participants.  While such research suggests that weight stigma in health care is undesirable and may have serious consequences, the full scope of the consequences of weight stigma remains underexplored. In particular, there is a lack of consideration in the weight stigma literature to the potential consequences of health care providers medicalizing overweight and obesity. Fat studies scholars provide some insight into this issue, arguing that weight-centred approaches to health may lead to weight cycling and disordered eating (O’Reilly & Sixsmith, 2012). In a qualitative interview-based study, Bombak et al. (2016) also argue that the language used to discuss ‘obesity’ in (reproductive) health care leads to clients feeling emotionally distressed. Despite this recent work, we still know little about the effects of medicalizing weight in health care. How do patients experience the medicalization of their bodies in health care settings? What are the implications of medicalizing weight on health care providers’ practices and behaviours? Given the increasing recognition of the high prevalence and consequences of weight stigma, researchers have begun to argue that we need to find ways to reduce such stigma, including within health sectors (Daníelsdóttir et al., 2010; Puhl & Heuer, 2009; Puhl & Heuer, 31  2010). D’Zurilla et al. (2004) identify four central aspects of social problem solving: problem definition and formulation; identifying potential solutions; decision making; and implementation and evaluation of solutions. Given the centrality of problem definition to problem solving, it follows that attempting to mitigate weight stigma in health care requires a thorough understanding of how weight stigma can be conceptualized. As already discussed, however, despite protestations in the literature that weight stigma is an urgent health and social justice issue that requires redressing, there is less conceptual clarity, which is thus a focus of my study. Next, I discuss what is known about reducing weight stigma.  2.4 What does current research say about weight stigma reduction? Here I focus specifically on what is known about how to reduce weight stigma and, in particular, on what has been learned through weight stigma reduction intervention studies, from which the most parallels can be drawn to my research. While actions, like legislation or policy changes, have been proposed to overcome weight discrimination (Puhl, Neumark-Sztainer, Austin, Suh &Wakefield, 2016; Puhl, Suh & Li, 2016), such considerations were out of scope of my dissertation.14  Perhaps as a result of limited understandings of what weight stigma is or what causes and perpetuates it, we know little about how to reduce the problem (PHSA, 2013a). Studies specific to weight stigma reduction are limited in number. As Daníelsdóttir et al. (2010) explain in their systematic review of the weight stigma reduction literature, there is surprisingly little research on the subject of weight bias reduction and even less evidence on what might actually work to                                                 14 Such considerations were out of scope because they did not fit within the boundaries of the case of focus, BalancedView, described later.  32  address this stigma. In their review of the literature, Daníelsdóttir et al. (2010) found only 16 weight stigma reduction intervention studies. Most of these were either ineffective in reducing weight stigmatizing attitudes or had methodological problems (e.g. utilized post-test only questionnaire designs, in which the effects of the intervention could not be ascertained). A more recent article using a meta-analysis found 30 weight bias intervention studies15 (including not only published articles but also graduate theses) and, after calculating effect size for each study, concluded that, on average, weight stigma intervention studies had a small but positive effect on weight-biased attitudes and beliefs (Lee et al., 2014). The relative lack of research into how to reduce weight stigma was also noted during this project. A literature review on weight stigma reduction, led by my partners at the PHSA as part of the BalancedView project reiterated this point. As they argue: “The bottom line . . . is that we really don’t know what’s going on, what works to decrease prejudice or why, in part because we still have a relatively poor understanding of prejudice” (PHSA, 2013a, p 28). They also noted that there remains a particular deficiency of research on reducing weight stigma in health care. Alberga et al. (2016) conducted a literature review to assess the state of weight stigma reduction interventions in health care arenas specifically. They located 17 intervention studies, with 15 focused on students in health disciplines and only two studies specific to practicing health care professions, illustrating an ongoing gap in knowledge on reducing weight stigma among practicing health care professionals.  Through reviewing the references of the systematic reviews by Daníelsdóttir et al. (2010), PHSA (2013a), Lee et al. (2014) and Alberga et al. (2016), as well as my own literature review, I                                                 15 Only included studies with a mean age of 18 years or older. 33  only found four studies specific to weight stigma reduction among current health care professionals, three of which were published in journals and one of which was a dissertation. Journal articles included: McVey et al. (2013); Gujral, Tea and Sheridan (2011); and Falker and Sledge (2011). Frick’s (2007) work was a dissertation that focused on current health care professionals and health students. From this small body of research on reducing weight stigma in health care, coupled with the more generalized research studies on weight stigma reduction, insights can be gleaned into possible strategies to reduce weight stigma. Both of the primary psychosocial explanations for weight stigma – (1) attribution theory and (2) social consensus theory – have been translated into and tested as weight stigma reduction interventions. Additionally, the concept of (3) using empathy (referred to as ‘evoking empathy’) to try and reduce weight stigma has also been tested (Lee et al., 2014), along with (4) contact theory, (5) consciousness raising and cognitive dissonance, (6) Health At Every Size® (HAES®) information (Health At Every Size and HAES are registered trademarks of the Association for Size Diversity and Health and used with permission) and (7) multi-strategy approaches. Each of these possible strategies and accordant intervention studies are discussed next, with specific attention to when the study pertains to health care providers. 2.4.1 Attribution theory and weight stigma reduction  Researchers examining attribution theory suggest that a possibly useful stigma reduction technique might be to raise awareness of the well-evidenced complexity of causal factors leading to obesity via information provision – pointing to factors such as the obesogenic environment or genetics – rather than allowing the focus to remain on lifestyle choices as the primary explanation. The hope is that if people learn that weight is much less within an individual’s 34  control than is often thought, stigmatizing attitudes will be lessened (Puhl & Brownell, 2003). The text boxes below summarize the extant studies using this approach. I begin with studies using this approach with youth. Studies using attribution theory with youth Anesbury and Tiggemann (2005), Bell and Morgan (2000) and DeJong (1980) have all explored whether providing information on the difficulties of controlling weight to children or youth reduces weight-biased attitudes. The results of these studies are mixed. Anesbury and Tiggeman’s (2005) study was ineffective at reducing negative stereotyping, despite modifying controllability beliefs. Bell and Morgan (2000) also had mixed findings, with older children who learned about genetic explanations for obesity rating ‘obese targets’ more negatively, whereas younger children learning the same information rated ‘obese targets’ more positively. DeJong’s (1980) study, divided youth into groups, with some provided with behavioural explanations for obesity and others provided with a medical explanation. The results were positive, those who learned about medical explanations scored as less biased.  Attribution theory has also been tested with adults in the general population, as per the studies described in the below text boxes.  Teachman et al.’s (2003) attributional intervention  Teachman, Gapinski, Brownell, Rawlins and Jeyaram (2003) utilized a post-test only design to explore both implicit and explicit attitudes16 among 144 adult pedestrians at a beach. Participants answered a questionnaire based on either no prime, after reading an article on genetic causes of obesity or after reading an article on behavioural causes. Participants took an Implicit Association Test (IAT) and the Fat Phobia Scale. Results were negative: those who read the article on genetics did not score any more positively on either implicit or explicit attitudinal measures.                                                    16 Implicit attitudes are more automatic and affective than explicit attitudes. Explicit attitudes are those aligned with our belief system that require consideration of the evidence.  35  Lippa and Sanderson’s (2012) attributional intervention  In this study, individuals over age 18 from the general population (N = 396) read brief vignettes about obesity, after which they took the short form of the Fat Phobia Scale. There were three experimental conditions: a genetic article, a genetic and environmental article and an environmental article. The study was ineffective. In conditions one and two, beliefs about the causes of obesity changed, but not weight bias. However, half of participants in all conditions read about the impact of diet and exercise on obesity. Those results were not the focus of the current paper, although they did say there was no difference in stigma between groups who read behavioural information versus those who did not. The interpretability of this study is limited by no measurement before the intervention, the use of a short text-based vignette and minimal differences between conditions.  Khan, Tarrant, Weston, Shah and Farrow’s (2017) attributional intervention  Participants (N = 463) from a general US adult population were divided into three groups. Participants were shown a photo of a man with obesity and were provided with one of three different accounts of the cause of his obesity: genetic, psychological or behavioural. After seeing the photo and receiving the explanation for his obesity, participants completed measures including an adapted version of the Fat Phobia Scale and a stereotypes checklist. While the Fat Phobia Scale normally assesses bias towards obese people in general, in this case it was reframed so that participants completed it based on their feelings about the male subject in the intervention. They also evaluated empathy towards the male subject. Further, to assess generalized weight bias in addition to weight bias specific to the male subject, they also used the Beliefs About Obese Persons Scale and the Anti-Fat Attitudes Scale. The results were mixed. The positive results were that those exposed to the information on genetic causes held the least subject specific bias towards the obese man, while those exposed to information on behavioural causes had the most bias towards the subject. Those exposed to genetic etiology information were also more likely to empathize with the obese individual. In contrast, however, they found no significant differences between groups on the measures of generalized weight bias. These result suggests that while brief information provision about obesity being out of an individual’s control may help decrease person specific bias, this alone will not help reduce more generalized weight bias towards about the broader population. Of note, this study also did not employ baseline measurement, making it difficult to assess if group differences were attributable to the intervention or other factors. In addition to studies examining the effects of attribution theory on adults in the general population, attribution theory has also been explored as a means of reducing weight bias among university students, as per the studies highlighted in the below text boxes.    36  Crandall’s (1994) attributional intervention  Crandall (1994) divided psychology students (N = 42) into two groups. One received a message about genetic causes of obesity, the other a message about the effects of stress on illness. They utilized a post-test research design, with the Anti-Fat Attitudes (AFA) Questionnaire.17  The results were positive. Participants who were informed of a genetic cause of obesity scored as less biased on two of the three subscales (willpower and dislike, but not fear of fat). However, the AFA has been critiqued due to the inclusion of the statement “I don’t have many friends that are fat” in the dislike subscale. This may be a measure of proximity (i.e. not having many fat people around), more so than dislike (Morrison & O’Connor, 1999).  Diedrichs and Barlow’s (2011) attributional intervention  Undergraduate psychology students (N = 85) were divided into three groups: 1) an intervention condition with a lecture that emphasized the multiple determinants of weight and discussed weight bias and size acceptance; 2) a comparison condition with a lecture on behavioural determinants of obesity; or 3) control with no lecture. Measurement occurred before and after the lectures and at three-week follow-up via the Anti-Fat Attitudes Test (AFAT).18 Results were mixed. After the intervention and at follow-up, the intervention group showed positive changes in the attractiveness and controllability subscales of the AFAT, but not the disparageability subscale. This may provide evidence in favour of attribution interventions, however, since the intervention condition also received information about weight bias and size acceptance it is questionable which (combination) of these factors the effect can be attributed to. Persky and Eccleston’s (2011) attributional intervention In this study, medical students (N = 110) were randomly assigned to either read a short scientific article about genetic causes of obesity, behavioural causes or about another topic altogether (headaches). They then participated in a simulated online virtual primary health care interaction with an ‘obese patient’19 in which the patient described having a rash, knee pain and intermittent shortness of breath, which students were told to react to. After the simulated interaction, participants took a questionnaire with the Obese Persons Trait Survey.20 No measures were used before the intervention. Relative to participants in the other conditions, those in the group that emphasized multiple causes of obesity scored as less biased on most (but not all) attitudinal questions.                                                    17 An often used, 13 question questionnaire with 3 subscales: dislike, fear of fat and willpower. 18 A questionnaire with three subscales that is different than the AFA. 19 They show an image of the obese patient simulated in the article, however, the patient, to my eyes, does not appear obese. The patient is an attractive, young, white female who is perhaps ‘curvy’, but not visually appearing obese.  20 This survey has 10 questions regarding traits, including laziness. 37  O’Brien, Puhl, Latner, Mir and Hunter’s (2010) attributional intervention  O’Brien et al. (2010) conducted similar attributional work with pre-service health students (N = 159). Students were divided into one of three tutorials: 1) a tutorial that emphasized the controllable causes of obesity (diet and exercise); 2) a tutorial that emphasized the uncontrollable causes of obesity (genetics and environment); or 3) a control tutorial focused on alcohol use. They measured explicit attitudes at baseline and after the intervention through the AFA. Implicit attitudes were measured through the IAT before and after the intervention.21 Results were mixed. The results were favourable on the implicit measures22 and mixed (negative on some dimensions and positive on others) on the explicit measures. Analysis of the explicit measures showed that although there were no significant differences across groups, there was a significant decrease on the dislike subscale scores in group two after the intervention. That said, scores from the willpower dimension of the AFA were rated as more stigmatizing in group two at the post-intervention measurement point relative to their pre-intervention scores.  Overall, attribution approaches when used alone are not the most promising. Some studies presented above were ineffective in mitigating biased attitudes or had mixed results. Of the studies that have found this strategy effective, there were often methodological problems calling into question the interpretability of the results, for instance not employing measures both before and after the intervention to assess for changes or direct effects. Perhaps attribution approaches to stigma reduction are not entirely successful due to the inherent complexity of the weight stigma issue. Deeply embedded social problems often have complex roots – as Farrell (2011) illustrated with respect to weight stigma – and solutions not responsive to this complexity are unlikely to be effective. Also of note, many of the interventions described above were brief in nature. It is possible that information provided during a brief intervention may be insufficient to leverage change.                                                  21 They also used a scale regarding beliefs about obese people to further assess whether perceptions of controllability changed.  22 The implicit attitudinal analysis showed a more favourable rating of fat people among those in group two in the post-test on two dimensions of the IAT (‘good/bad’ and ‘motivated/lazy’) relative to the other groups and to the pre-test. 38  2.4.2 Social consensus theory and weight stigma reduction  In addition to the above attribution approaches to stigma reduction, social norms/consensus theory has also been examined as a weight stigma reduction strategy, albeit in fewer studies. Stigma reduction, according to social norm/consensus theory, may be achieved through exposing people to the positive attitudes of others, especially if those others are influential members of society. Studies using this approach are described in the text boxes below.  Puhl, Schwartz and Brownell’s (2005) social norms/consensus study  Puhl et al. (2005) conducted three experiments that explored social norms/consensus theories with American university students with positive outcomes. In experiment one, students (N = 60) took a pre-test questionnaire about their explicit beliefs and attitudes about obese people. One week after this, students returned to take a post-test. Prior to beginning the post-test, however, students were presented with (manufactured) data about how their peers had rated obese people in the pre-test. Students were either told that others rated obesity more or less favourably than them. Learning about the positive ‘norms’ among their peers improved attitudes at the post-test. In the second study, 55 students similarly took a pre-test and a week later returned for a post-test. Before the post-test they were informed of their (manufactured) peers’ responses. Half the group received information about peer norms from a community college and half from an Ivy League university. The results showed that the in-group norms – the Ivy League norms – were more influential on participants than community college norms. In the third experiment (N = 200), a social norm approach was compared with other means of mitigating weight stigmatizing attitudes. Following the pre-test and prior to the post-test students were presented with one of the following: 1) manufactured Ivy League norms; 2) favourable information about obese people framed as ‘scientific truth’; 3) a vignette on the challenges of controlling obesity; 4) a vignette about obesity as within personal control; or 5) no additional information. Participant associations of obese people with positive traits increased and associations with negative traits decreased in groups one and two in the post-test. Negative associations decreased in group three. The intervention for group two – presentation of information framed as ‘factual’ and ‘scientific’ – was most influential. A potential issue with this study relates to response bias. If participants believe their peers view obesity more favourably than them they may be inclined to change their responses, but the extent to which this represents a true shift in beliefs or attitudes is unknown.     39  Zitek and Hebl’s (2007) social norm study  Zitek and Hebl (2007) also used a social norm approach with 270 women at an American University. They listened to statements about prejudice towards five groups (one of which was people with obesity23), in which the speaker either condoned prejudice, condemned prejudice or did not provide verbal information on whether they condoned/condemned discrimination (control). There was a post-test and follow-up questionnaire (no pre-test measures). Results were positive. At both post-test and follow-up, students who listened to the speaker either condemn or condone discrimination were more likely to respond similarly, compared to the control, as based on questions about whether they agreed or disagreed with statements about prejudice and discrimination being appropriate or inappropriate.  Gumble’s (2012) social consensus study  This study was a dissertation that tested the effect of a social consensus intervention on 110 American university students. Individuals filled in a questionnaire with explicit bias measures (the Obese Persons Trait Survey) as well as the Implicit Association Test. A week later they returned and were provided with fabricated results from other individuals (with weight status manipulated) who supposedly filled in the same measures. They then repeated the questionnaires. The author found that social consensus impacted levels of explicit bias. Decreases to implicit bias were also seen when the fabricated information was from in-group members (based on weight status).  As the above text boxes shows, the results from these studies are promising with respect to mitigating weight-biased attitudes. However, more research is required to develop further insights about the effectiveness of social consensus approaches to weight stigma reduction. 2.4.3 Evoking empathy and weight stigma reduction  In addition to interventions designed around attribution or social norm/consensus theories, other approaches have also been proposed to reduce weight stigma. One such strategy is that of ‘evoking empathy’ – a strategy that has reportedly had some success in stigma reduction relating to other persons, such as those living with AIDS (Gapinski, Schwartz & Brownell, 2006). This strategy suggests that through promoting an understanding of the challenges of living with a stigmatized condition, compassion and empathy will be fostered and stigma decreased.                                                 23 The others were black people, gay people, racist people and ex-convicts. 40  Regarding weight stigma, Thomas (2016) has suggested that social empathy is the remedy to stigma and that the two cannot exist simultaneously. To date, studies using this approach have had mixed results, as shown in the text boxes below, beginning with interventions specific to youth.  Studies using an empathy approach with youth  Hennings, Hilbert, Thomas, Siegfried and Rief (as cited in Daníelsdóttir et al., 2010) Hennings et al. (as cited in Daníelsdóttir et al., 2010)24 provided a 20-minute video of overweight youth to high school students (N = 602). The video described experiences with weight-based discrimination. A 13-item researcher-developed questionnaire was utilized and administered pre- and post-test. The outcomes were reportedly negative. Anti-fat prejudice increased following the intervention, although there was an enhanced understanding of the challenges of being obese.   Irving (2000) Irving (2000) also used an empathy approach. Irving aimed to educate elementary school youth  (N = 45) that bodies come in a variety of shapes and that teasing about weight hurts people’s feelings. This was done through a puppet show with a size acceptance, anti-diet, no teasing theme. For measurement, some participants took a pre-test questionnaire only. The rest took a post-test questionnaire only.25 The questionnaire asked students to rate different body shapes (heavy, average and thin) on six dimensions. Results were mixed. Students in the pre-test rated heavy people as worse than average weight people on all six dimensions. Students in the post-test rated heavy people as worse than average people on three of six dimensions (lazy, sad and stupid), but the same on the other dimensions (friends, teased and cute). Given that different participants took the pre-test than the post-test the interpretability of the results is limited.                                                    24 I could not access this 2007 article, as it was not in English. I report on it based on the summary in Daníelsdóttir et al. (2010). 25 Images/drawings of large, medium and thin size children were each rated on six bipolar adjectives (cute/ugly, lazy/works hard, teased/not teased, friends/no friends, stupid/not stupid and happy/sad), as adapted from the Figure Rating Scale. 41  Rukavina, Li and Rowell’s (2008) empathy intervention with university students Rukavina et al. (2008) developed a six-week intervention for undergraduate kinesiology students that utilized an empathy approach (N = 95). Students participated in lectures designed to raise awareness of the challenges of being obese and also participated in experiential learning through conducting physical fitness tests with elementary youth. Pre- and post-test measures of explicit biased attitudes were collected via the AFAT and two semantic differential scales (stupid/smart, lazy/motivated). The results were mostly negative. There was no change in the semantic differentiation scales. There was improvement on the weight control/blame subscale of AFAT, but no statistically significant change on the other two subscales.  Falker and Sledge’s (2011) empathy intervention with health professionals This study focused on evaluating the effectiveness of a web-based bariatric sensitivity training educational module with health care staff. Methods included a survey on attitudes towards bariatric patients before the module and another survey one month after completing the module. The text-based sensitivity training aimed to educate about the many causes of obesity and to increase sensitivity through enhancing knowledge/understandings of obesity. The underlying theory of the course was that education and knowledge would increase empathy and decrease stigma. A total of 600 individuals were exposed to recruitment information. Only 101 individuals completed the pre module surveys and 30 staff completed both the pre-module survey and post-module survey. Eighty percent were nurses, with the rest compromising health care technicians and unit secretaries. Eighty seven percent were female. The primary measure was the Care of the Bariatric Patient Nursing Survey, which was created to measure stereotypical attitudes before and after the course. The survey had eight statements and a 4-point likert agreement scale. Statements were focused around things like sensitivity to the needs of patients and awareness of how attitudes can impact care. The mean scores lowered at the post-module measurement point, indicating an improvement in sensitivity and a positive effect. A limitation to this study is that the construct of attitudes was not well measured through the survey created. Thus, while sensitivity may have improved, we know less about the effects of the intervention on attitudes (or for that matter, behavioural manifestations of weight bias). There was also a low completion rate (approximately 30 percent) and a primarily nursing-based, female demographic.    42  Gloor and Puhl’s (2016) empathy and perspective taking study with adults  Gloor and Puhl (2016) tested the effects of four different stigma reduction interventions with American adults (N = 650). Participants were assigned to either a control condition or one of four experimental conditions: 1) empathy; 2) perspective taking; 3) causal information; or 4) a hybrid of empathy and causal information. In the empathy intervention participants read a first person narrative about a man with obesity and the challenges he faced trying to lose weight. The narrative emphasized contributors to his obesity including personal decisions and factors outside his control. In the perspective taking condition participants had to write about what they imagined a day in the life would look like for a heavier individual. Participants in the causal group read information about the complex causes of obesity, including those outside a person’s control. Those in the hybrid condition read condensed text from the empathy and causal interventions. Measures were taken after the intervention and included the short version of the Fat Phobia Scale, six empathy questions and six affective reaction questions (including negative affect and sympathy or concern), among other measures. The results were mixed (improved empathy but no impact on fat phobia relative to control). Those in the empathy or perspective taking conditions had more empathy and improved emotional responses to people with obesity than those in the other groups. However, no experimental group had decreased fat phobia compared to the control. The direct effects of the empathy and perspective taking experiments on weight bias itself are difficult to ascertain due to no measurement occurring before the experiments. The experiments in this study were also all very brief. It is possible that a more in-depth intervention would have had an effect on weight-biased attitudes.  Gujral, Tea and Sheridan’s (2011) empathy intervention with health professionals Gujral et al. (2011) conducted an online survey (N = 332) of nurses’ attitudes (measured by ATOP) and beliefs (measured by BAOP) from two different hospitals. One hospital provided annual bariatric sensitivity training and one did not. No statistically significant differences were found between hospitals, however the authors note that the study lacked power. The ATOP scores were slightly better among those who received sensitivity training, but the BAOP scores were similar across conditions. The authors concluded that sensitivity training may slightly improve attitudes, but not beliefs. These results held regardless of participant BMI. This study was limited as the sensitivity training was not well described and its content largely unknown, thus it is difficult to draw any theoretical insights from the study with respect to what strategies may or may not improve weight bias. All they said about the training was that it was web-based, overviews obesity, stigma and discrimination, and relevant equipment and resources. It was unclear what the guiding stigma reduction principles were. I have put this study with the empathy interventions as it claimed to be a sensitivity training, however, a more accurate depiction may be that the stigma reduction mechanism underpinning this study was unclear.    43  Cotugna and Mallick’s (2010) empathy and attribution intervention Cotugna and Mallick (2010) put dietetics and health promotion students on a diet for one week (1200 calories for females, 1500 for males). During the week, students had to journal about their experiences. While the ethics of this methodology and type of intervention can be called into question, the study was effective. They found significant reductions in explicit fat phobic attitudes as measured by the Fat Phobia Scale taken before and after the intervention. The journal entries also showed that students had a newfound empathy for the challenges of dieting and weight loss. Perhaps this study was more effective because it promoted empathy more so than sympathy, per se, and addressed the challenges of controllability through experiential learning. Regardless, given that this study was to do with pre-service health care providers, it is relevant to my study.  In addition to these intervention studies, Meadows et al. (2017) explored the relationship between medical school experiences (including perspective taking or ‘empathy’ training) and weight-biased attitudes towards patients with higher weights. Using a large data set of 3,576 students in 49 different US medical schools they found that, based on one question that asked students to rate the number of hours they spent in medical school participating in training on seeing things from a patient’s perspective, perspective-taking training was associated with only small decreases in biased attitudes towards heavier patients.  Overall, empathy interventions and approaches have shown mixed results in terms of reducing stigma. When used as a stand-alone strategy, empathic interventions may lead to the paradoxical effect of increasing rather than reducing weight stigma (Daníelsdóttir et al., 2010). Perhaps this is because increasing awareness of the difficulties of living in a fat body, without simultaneously addressing blaming attitudes that fatness is controllable, worsens antipathy towards heavier people. Regardless, one important consideration is to clarify as much as is possible, what is meant by ‘empathy’, something that has had little attention in the weight stigma literature so far given that this is a new area of inquiry. How empathy has been defined within the weight stigma literature is discussed next.  44  Gloor and Puhl (2016), in an article on weight stigma reduction, defined empathy as the reactions a person has to the observed experiences of another person. They state that the ability to take the perspective of another is often conceptualized as central to empathy. Gapinski et al. (2006) also perceive perspective taking as key to empathy. Gloor and Puhl (2016) further note that there may be cognitive and emotional elements to empathy. This is also taken up by Meadows et al. (2017) who operationalize empathy as having a cognitive component (i.e. ‘perspective taking’) and an affective component (i.e. ‘emotional empathy’).  Other literature also provides some insight into the empathy concept, discussed next, before continuing to discuss what is known about weight stigma reduction specifically. Rogers (1957) defined empathy as: “the capability to sense the client's private world as if it were your own” (p 99). Richardson, Percy and Hughes (2015), in a discursive review of health education literature articulate that empathy is often poorly defined in scientific studies. They also suggest that empathy, caring, kindness, sensitivity and compassion are interrelated concepts that make it challenging to ensure we are referring to the same thing. In distinguishing between these ideas they state, “Unlike caring and compassion, empathy is mostly designated as a cognitive or emotional concept and is less likely to be described in terms of behaviours” (p 2). Further they argue, “Despite the complications arising from the vagueness of the definitions of caring, compassion and empathy, it is clear that health service users can detect these and other related qualities in nurses’ behaviours and attitudes” (p 2). This suggests that empathy is something emotionally understood more so than understood in a traditionally scientific sense. Their articulations also raise the question of whether empathy is what needs to be generated among health care providers to reduce stigma or whether concepts like compassion are also needed. In 45  distinguishing between the concepts of compassion and empathy, Singer and Klimecki (2014) in a neuroscience article conceptualize empathy as feeling the emotions of others, including distress, and compassion as a related notion of wanting to care for others. As such, they see empathy as likely to lead to a ‘pulling away’ effect due to empathic distress and compassion as more pro-social, again raising the question of what should be the target of emotion related interventions: compassion or empathy? Overall, it is clear that empathy (and/or compassion) and weight stigma reduction require further study, as so far this is a new and emergent area. Contact theory approaches to stigma reduction are discussed next.  2.4.4 Contact theory approaches to weight stigma reduction  Proponents of contact theory suggest that stigma can be reduced through exposure to or contact with stigmatized groups. The hypothesis is that contact with stigmatized people allows the non-stigmatized an opportunity to counter their stereotypical beliefs about those with a particular stigmatized attribute. In other words, contact may result in a disintegration of their prejudicial categorizations about particular groups. Contact theory has been used to successfully reduce stigma in other areas. For example, see Couture and Penn (2003), regarding mental illness.  While not an intervention study, Meadows et al. (2017) explored the relationship between medical school experiences (including favourable contact with obese people) and weight-biased attitudes towards patients with higher weight. Based upon the responses of 3,576 students in 49 different US medical schools they found that favourable contact or interactions with heavier patients were significantly associated with less biased attitudes. However, only intervention study I know of utilized a contact-based intervention to try and reduce weight stigma, as described in the below text box. 46  Roberts et al.’s (2011) contact study  Roberts et al. (2011) paired third year medical students (N = 4) with patients with obesity who were undergoing evaluation for bariatric surgery for a one-year period. Students took what seemed to be a researcher-developed pre- and post-test attitudinal questionnaire26 and wrote regularly in a journal about their own stereotypes. Their attitudes were compared to several peers who did not participate. Results were mixed. The questionnaire showed that some measures improved (e.g. empathy), while others stayed the same (e.g. perceptions of willpower). Depending on interpretations, some measures could also be considered to have worsened (e.g. students were more likely to see obesity as a chronic disease for which they would recommend weight loss surgery). The journals showed that students were able to challenge their stereotypes as a result of extended interactions and relationships with the patients. This study had a very small sample size that limits its generalizability.  Regardless of how the results of the above Roberts et al. (2011) study can be interpreted, a critical issue with contact theories of stigma reduction is that such theories cannot account for why weight stigma is an increasing problem in a culture where rates of overweight and obesity have also seemingly increased. Another challenge with this strategy, as used in the above study, is that the mechanisms within it that may reduce stigma are unclear (PHSA, 2013a). Contact could have been effective because participants were exposed to counter-stereotypical information, or could it have been for some other reason, perhaps because contact promotes relationship building and, thus, empathy and compassion?  2.4.5 Cognitive dissonance or consciousness raising  Consciousness raising is another strategy that has been suggested to attempt to reduce stigma. This strategy is based on encouraging critical reflection about one’s values, attitudes or behaviours. The idea is that critical reflection may work to reduce stigma as it provides an opportunity to create cognitive dissonance insofar as individuals may experience discomfort when they realize their attitudes or behaviours might not be consistent with their overarching                                                 26 Only limited details on the questionnaire were provided. 47  values. This discomfort is thought to increase the likelihood that individuals will modify their attitudes and behaviours to suit their value systems (PHSA, 2013a). Ciao and Latner (2011) tested this approach for its utility to weight stigma reduction and compared it to social consensus. Their study is summarized below.  Ciao and Latner’s  (2011) cognitive dissonance study   Ciao and Latner (2011) compared cognitive dissonance to social consensus means of weight stigma reduction. Participants were college undergraduates in psychology (N = 64). A pre- and post-test questionnaire design was used. One week after taking the pre-test (based on the AFAT), participants in the cognitive dissonance group (n = 21) were informed that their AFAT scores were not consistent with values of equality and fairness and were asked to immediately take the post-test. The results were positive. The cognitive dissonance group scored more positively in comparison to pre-test and to the social consensus group (the social consensus intervention, which entailed providing information about how peers had responded, failed).  While the Ciao and Latner (2011) study suggests cognitive dissonance may be helpful to weave into weight stigma reduction interventions, this approach is only just beginning to be explored as a weight stigma reduction strategy, which makes commenting on its potential effectiveness challenging. Also, while the above study was successful in improving weight stigma attitudinal scores, it is not clear that attitudinal change necessarily relates to behaviour change. An overall shortcoming with most of the extant weight stigma reduction research is the use of attitudinal measures of weight stigma without a corresponding behavioural measure. My study addresses this gap and measures not only weight-biased attitudes before and after the BalancedView intervention, but also self-reported behaviour and practice changes among participating health care providers.  48  2.4.6 Multi-method interventions  Some studies on mitigating weight bias have tested a multi-strategy approach based on some combination of the above strategies, sometimes with other components woven in. As the below intervention studies highlight, multi-prong strategies show promise.   Swift et al.’s (2013) multi-method intervention  Swift et al. (2013) tested a multi-strategy approach to stigma reduction with 19 graduate nutrition students and 24 third year medical science students. Students were shown two 17-minute videos on weight stigma. The video included an empathy component, an attribution component and counter-stereotype information. Explicit and implicit measures of attitudes were taken pre- and post-test and at follow-up.27 Results were mostly positive: explicit attitudes improved, however implicit attitudes were more resistant to change.  Robinson, Bacon and O’Reilly’s (1993) multi-method intervention This study utilized a three-prong information provision strategy that emphasized attribution theory, the inherent beauty in all body sizes and provided information on weight prejudice with a convenience sample of 40 white American adult women. The intervention involved group sessions and individual sessions. Pre- and post-test measurement occurred using the long version of the Fat Phobia Scale. The study was effective: attitudes improved after the intervention.  Hague and White’s (2005) multi-method intervention This study focused on teachers and student teachers (N = 258). Participants participated in an Internet module that presented information on weight stigma and the multiple causes of obesity and emphasized a non-diet, health-centred approach to weight. Although all information provided was the same, it was provided in different ways to test social consensus. Some participants received the information with no image or credentials attached, others with presenter credentials (PhD and Registered Dietitian) but no image and others with an image of a non-obese presenter with credentials or an obese presenter with credentials. Measurement occurred via a questionnaire at pre- and post-test and follow-up.28 Reductions in weight-biased attitudes occurred and were sustained at follow-up. Those exposed to the ‘credible’ obese presenter showed the greatest improvement.                                                  27 Via the Fat Phobia Scale, Beliefs About Obese People Scale and the dislike and willpower subscales of the AFA Questionnaire. The IAT was used to measure implicit attitudes. 28 They used the AFAT and questions assessing trustworthiness of information at pre-test, post-test and 6-week follow-up. 49  Gapinski et al.’s (2006) multi-method intervention Gapinski et al. (2006) is the exception to the success found in above three multi-strategy approaches. The authors utilized a twofold-approach (N = 108): evoking empathy and countering stereotypes in a controlled experiment. American female university students in the experimental group watched a set of two short videos portraying an overweight individual. The first video was about the challenges of being overweight, the cruel treatment often faced and subsequent feelings of hopelessness (the control group watched an unrelated video). The second video was a clip with stereotypical or counter-stereotypical information about obese people (e.g. motivated/unmotivated, smart/not smart, professional/unprofessional). The research used a post-test only questionnaire design. Measures included the IAT, several questionnaire-based measures of explicit attitudes and an empathy scale.29 The results were negative: there were no differences between groups on implicit or explicit measures.  Hilbert’s (2016) multi-method intervention This article focused on a multi-component intervention used in two studies that drew largely on attribution theory and included other components. Each study tested the effects of a brief, interactive, computerized, multi-component intervention with adults. The intervention was a 60-minute computerized course that provided an opportunity to reflect on beliefs, led participants through guided imagery and provided information challenging controllability beliefs (e.g. genetic and environmental contributors to body weight). It also discussed cultural pressures to be slim, how beauty norms differ across cultures and research on weight prejudice and its consequences. Study one (N = 128) was with university students. Half the participants were in an intervention group and the other half in a control group. For the intervention group measures were taken at baseline and 10 to 16 days after the course. Measures included the Anti-Fat Attitudes Test (AFAT), Beliefs About Obese Persons Scale (BAOP) and an online IAT. The results included significant improvements (medium effect) on the AFAT right after the course for the intervention group compared to control. Controllability beliefs (as measured through BAOP) decreased significantly (large effect). There was no difference between the control and intervention group on implicit attitudes. Study two (N = 128) was conducted with the general population, not students (63 in the experiment group and 65 in the control). Measures included the Anti-Fat Attitudes Test, BAOP and the IAT, among other measures at baseline and follow-up (four weeks later). While study one had a positive effect, the second study did not show any significant effect on explicit attitudes or controllability beliefs, except for among those with higher education levels. The author concludes that the intervention was useful in the short-term (study one, university sample) and in the long-term for those with higher education.                                                   29 Measures of explicit bias included: semantic differential scales regarding thin and fat people (motivated, unmotivated, etc.); a feelings thermometer to assess feelings towards thin, average weight or obese persons; a preferences questionnaire; and resume ratings. 50  Overall, as the studies in the text boxes above illustrate, the evidence is encouraging, albeit modest, for interventions that employ multiple strategies (Daníelsdóttir et al., 2010; PHSA, 2013a). While the Gapinski et al. (2006) study was not effective, it is worth noting that this study did not include any form of measurement before the intervention, thus, it is challenging to ascertain the effects of the intervention. The other studies, which did employ measurement both before and after the intervention, were promising, although the intervention by Hilbert (2016) was not effective in the longer term for those with lower education levels.  Given multi-component interventions are more likely to be responsive to the multiple and complex contributing factors to stigma (Link & Phelan, 2001) they thus may be useful to incorporate into weight stigma reduction. Multi-method approaches also lend well to longer, more in-depth interventions, which may be more helpful in mitigating stigma than very brief interventions. Most of the interventions described above are only several minutes in length. Moving beyond very brief information provision was identified as helpful in producing longer term positive effects on weight-biased attitudes in one study (Swift et al., 2013). Currently, however, it is unknown what components of multi-strategy interventions are most effective. For example, Hilbert (2016) concludes that their study provides some evidence in favour of attribution approaches with educated populations, both in the short-term and longer term. However, the intervention had multiple components and the authors were unable to assess for the effects of the different components, thus it is unknown what aspects were most useful.  Another effective, multi-method intervention study, not discussed in the above table, was conducted by McVey et al. (2013). This study is worth discussing in its own right since it was specific to public health professionals in Ontario and was one of the four studies with health professionals that I found. The study is summarized below.  51  McVey et al.’s (2013) multi-method intervention for health professionals McVey et al. (2013) held a daylong forum for public health professionals in Ontario (N = 342) to reduce weight stigma.  The forum aimed to raise awareness of: weight stigma and its health consequences; ways to balance weight and health messages to avoid triggering disordered eating and body image concerns; and incorporating mental health promotion language into healthy weight messaging (i.e. focusing on mental health in addition to physical). A questionnaire was administered before and after the intervention and at six-week follow-up.30 A total of 42 participants took part in semi-structured, follow-up interviews aimed at understanding changes made since the intervention. Anti-fat attitudes were reduced after the forum and at follow-up, though the effect lessened over time. Similar findings were present for self-efficacy to address weight-related norms and internalization of media stereotypes about weight and appearance. Despite the promise of this study, the authors acknowledge it had limitations, including no control group.  What is particularly interesting about the study by McVey et al. (2013) is their unique emphasis on balancing weight and health messages to avoid triggering disordered eating and body image concerns. With this, the authors specifically strove to provide information on the harms of focusing on weight loss as a measure of health and emphasized disordered eating as likely consequence of focusing on weight. I suspect that a key aspect of the success McVey has had, both in this study and politically in Canada as a champion of the need to reduce weight stigma in health care, is due to her unique positioning on messages about weight and health. Health professionals are very focused on health as a desirable outcome. Thus, McVey et al.’s (2013) discussion of eating disorders as a consequence of weight-centred approaches to health may have been a viable way of creating buy-in among health professional participants for the notion that overly focusing on weight in health sectors is not always helpful.                                                  30 Measures included: the 13-item AFA Questionnaire; the Sociocultural Attitudes Towards Appearance Questionnaire (a measure of internalization of weight and appearance norms); a version of the Body Satisfaction Scale; and Self-Efficacy to Change weight-related social norms. 52  2.4.7 Health At Every Size and its potential for stigma reduction  The notion that focusing on weight may be stigmatizing and that taking the emphasis off weight may thus help reduce weight stigma was an important focus of my study. I discuss this idea below in relation to Health At Every Size (HAES) principles. Promoting HAES has been suggested by fat studies scholars as a way to help reduce weight stigma (McMichael, 2013; O’Reilly & Sixsmith, 2012). HAES is a wellness-centred approach to health that advocates taking the focus off weight and instead concentrating on enjoyable physical activity and healthy eating for all bodies (Association for Size Diversity and Health, 2014). It can be adopted at the individual level or promoted in health sectors. A HAES-oriented approach has been tested and compared to weight-centred, diet-focused approaches in randomized control trials and found to be more effective at improving individual psychological and physical health outcomes than weight loss-focused approaches (Bacon & Aphramor, 2011). Two important aspects of HAES are that weight does not necessarily relate to health, and that focusing on weight is harmful to health. Unlike the above mentioned stigma reduction strategies, with the exception of McVey et al. (2013) who loosely draw on ideas about HAES in their weight stigma reduction intervention, I only located one stigma reduction intervention that formally drew on Health At Every Size.31 This study is described below.                                                    31 This study occurred prior to HAES becoming a registered trademark of the Association for Size Diversity and Health. 53  Frick’s (2007) use of Health At Every Size concepts in a weight stigma intervention  As part of a dissertation, a one-hour in-service (N = 37) was provided for current health care providers or pre-service (nursing) health students that drew on concepts related to Health At Every Size and the trans theoretical model (TTM) of change. The TTM is based on the idea that change is a process that occurs over time and is fluid; change does not occur all at once. Stages of change include pre-contemplation, contemplation, preparation, action and maintenance. Measures included the Anti-Fat Attitudes Test and the Implicit Association Test. Measures were taken before and after the in-service and at four-week follow-up. Positive changes were seen after the intervention and sustained, though somewhat decreased, at follow-up.   The above dissertation by Frick (2007) suggests that incorporating HAES-related elements into a weight bias reduction intervention for health care providers and/or health students may be effective. The Hague and White (2005) study described above also is suggestive of this. Hague and White’s (2005) study drew on the related ideas of a non-diet, health-centred approach and was effective.  2.5 Other insights into weight stigma reduction  Beyond the theories tested in the stigma reduction intervention studies described above, insights concerning how to potentially reduce weight stigma can also be gleaned from the work of those in fat studies and critical weight studies, as well as from those engaged in applied approaches to weight stigma reduction. For instance, Norman and Petherick (2016) discuss how, in attempting to disrupt the dominant obesity discourse in a kinesiology classroom setting, it is useful to encourage students to think critically and evaluate the validity of dominant obesity messaging. To do this they purposefully introduce learners to the contradictions and inconsistencies underpinning the obesity epidemic discourse. Upon realizing the competing knowledge claims about obesity, students are forced into a position of critically grappling with the evidence. Bhagat and Jette (2016) similarly argue that for kinesiology students it is important to promote critical thinking by showing the contradictory research on obesity and its relation to 54  health and the evidence on the harms of the weight-based paradigm. The contention that interrupting dominant, medicalized understandings about obesity can be undertaken through encouraging critical thinking and exposure to multiple knowledge claims is taken up in this study.  Further insight can also be gleaned through Saguy (2013), who via a series of experiments shows that medical frames about weight relate to weight-biased attitudes, which suggests that targeting medical frames may be useful to weight stigma reduction. But what can we do with this information that the medicalization of weight may contribute to the problem? Beyond the possible use of HAES principles, how do we translate Saguy’s (2013) contention into a viable and testable intervention? I have taken this issue up in my study.  The use of drama has also been suggested as a strategy for addressing weight stigma (Lea & O’Reilly, 2016), including among health professionals. For instance, Kirk et al. (2014) discuss how, following a multilevel qualitative study on obesity management, they created a dramatic presentation of the relationship between a heavier individual and a health professional. The presentation is being used as an educational tool and preliminary pilot data show it has benefit in raising awareness among health care providers of tensions that may exist in such encounters (Kirk et al., 2014).32 Although investigating the role of drama to weight stigma reduction is not an explicit aim of this study, I do draw on and explore the utility of filmed re-enactments of health care encounters created for BalancedView as part of unlearning weight stigma.                                                   32 A video of the dramatic adaptation is available at on YouTube, see (Kirk, Price, Aston & Vallis, 2013). 55  Recently, building upon the dramatic work mentioned above, Kirk (2015a) released a time-limited online course on weight bias through the Canvas Network (2015).33 The course was free and publicly accessible for a limited time. The course:  . . . [p]rovide[s] an overview of the causes and consequences of obesity, with a focus on promoting a balanced understanding of the complex factors that have led to a rise in obesity rates globally and their implications for obesity management and prevention. As participants move through the content, they will be challenged to reflect on their own attitudes towards obesity and to critically appraise how these are shaped by broader societal attitudes. After completing the course, participants will have gained an appreciation of the causes and consequences of obesity and better insight into how to approach individuals experiencing obesity in a respectful and non-judgmental manner. (Canvas Network, 2015) The course aimed to provide participants with a better understanding of the complex causes of obesity and to learn how to approach people with obesity in a non-stigmatizing, respectful way (Kirk, 2015b). This latter aim is something that is also taken up in my study in that I explore how to not only reduce weight-biased attitudes – as the majority of intervention studies have done – but also how to encourage health care providers to ensure their practices are not biased.  Other pertinent weight stigma reduction resources have become available in recent years that also aim to address weight stigma in health care. For instance, The Binge Eating Disorder                                                 33 The course had five modules: one on exploring one’s own biases; a second on understanding the complexity of obesity as a health and social issue; a third on understanding weight stigma and where it comes from; a fourth on how to address weight bias and stigma; and a final module on best practices (Canvas Network, 2015). 56  Association (BEDA) has developed PDFs for various health care providers that provide basic information on weight bias.34 One of the strategies suggested in the guide to reduce weight stigma is using a health-focused rather than weight-centred approach to treatment. This approach is explored in my study for its utility in assisting weight stigma reduction. In addition to BEDA resources, another free online weight stigma course was developed in 2015. In November of 2015, the Rudd Center launched a publicly available course of weight bias and stigma.35 The Rudd Center’s University of Connecticut page stated the following of the course:  The course equips clinicians with strategies to improve provider-patient communication, make positive changes in the medical office environment, and increase awareness of personal biases that could unintentionally compromise patient care. (University of Connecticut, 2015) As with the course described above by Kirk (2015a, 2015b), this course strove to help participants learn to approach heavier people in a non-stigmatizing way. Learning objectives for this course were quoted as follows:   1) Recognize the sources of weight bias and stigmatization in health care settings. 2) Describe the adverse consequences of weight stigma on patients’ emotional and physical health. 3) Identify personal assumptions about obesity and body weight, and how these views can influence patient care.                                                 34 As an example, BEDA has a guide available titled, “weight stigma in the nutritional counseling setting: Guidance for professionals.” The purpose of the guide is to “. . . identify how weight stigma can impact your perception as a clinician, and provide concrete strategies to help you combat weight stigma in your nutrition counseling work” (BEDA, nd, p 1). 35 Available at improveobesitycare.org or http://ruddcentercme.org (UConn Rudd Center, 2015). 57  4) Improve communication skills to facilitate productive discussions with patients about weight-related health. 5) Implement clinical strategies to help patients with obesity set appropriate goals for lifestyle behavior change. 6) Identify strategies to improve accessibility and comfort for patients with obesity in the medical office environment. 7) Educate medical students about weight bias and provide resources for further training on this topic. (University of Connecticut, 2015) Of benefit, the above course appears to draw on more than one stigma reduction approach. It focuses on improving communication with patients, for example, as well as raising awareness of personal assumptions, both of which I explore in this study.36  2.6 Gaps in the literature: What do we need to know more about? While it has often been argued that weight stigma is a problem among health care providers, as I have shown in the preceding discussion, there has been less attention in the literature to efforts to conceptualize precisely what the problem is that we are trying to reduce. While studies over the last few decades have revealed much about weight-biased attitudes, knowledge on what constitutes weight stigma, in health care or otherwise, remains incomplete. More research is needed to build knowledge on what attitudes are considered weight-biased among health care providers. We similarly need to build knowledge on what practices are considered stigmatizing. We also need to debate further how the medicalization of weight relates                                                 36  Evaluation results from this course are unknown at this time, as far as I am aware. 58  to weight stigma in health care – from the perspectives of patients, providers and academics – as this is an area where there has been inadequate attention and extensive disagreement. Attention is also needed to understand how the concept of weight stigma relates to other emergent stigma theories, including, for example Link and colleagues’ (2004) notion of stigma as it relates to emotions. The importance of this problem definition work cannot be understated. As D’Zurilla et al. (2004) argue, problem identification and conceptualization are central to the social problem solving process. It follows that mitigating weight stigma in health care necessitates a simultaneous reflection on different theoretical perspectives that help to understand the problem. We also need to consider the perspectives of health care providers themselves. Understanding how health care providers think about and define weight stigma is an essential prerequisite to learning how to mitigate stigma among this population. More information is urgently needed about how to reduce weight stigma in health care, given the health and ethical implications of this stigma. Currently, we are lacking in both studies and knowledge on how to successfully reduce weight stigma in health settings. Further, few weight bias interventions have focused on changing the behavioural manifestation of weight bias (Lee et al., 2014), another area where attention is needed.  Relatedly, while recent studies have begun to test different approaches to try and reduce weight stigma, as articulated within this literature review, more attention is needed to the process of delivering weight stigma interventions. Is there a particular length of intervention necessary? Is text-based material sufficient, or would verbal or some other medium be preferable? What about the difference between online and in-person strategies to reduce weight stigma? Although web-based training for health care providers has been found to be effective in a general sense, 59  when compared to in-person learning (Doorenbos et al., 2010), we know little about the applicability of online interventions specific to health care providers and weight stigma.  Given these gaps, I aimed to build theory on how weight stigma can be conceptualized and show what might be done to reduce it (both attitudinally and behaviourally), in the context of health care. This aim was explored through a case study of BalancedView, an online intervention on weight stigma for health care providers in BC that was developed through a multi-year, collaborative, in-person process. As I describe more fully in the methodology chapter that follows, I used this case to explore the following research questions (RQs):  • RQ1: What are the different ways that weight stigma in health care can be conceptualized? • RQ2: What strategies can be employed to reduce weight stigma among health care providers in BC? The rationale for these questions, the case of focus and other methodological details are described next.   60  Chapter 3: Methodology  In this chapter I discuss my methodology. First, I remind readers of my research questions, articulate my case study research approach and describe the principles of a pragmatic paradigm that informed my work. I then describe the case of focus, BalancedView, which is an online course for health care providers in BC that aims to reduce weight stigma. I discuss the BalancedView case before my data collection methods in order to provide a frame of reference for the different methods employed at the various stages of the case study. I then discuss my methods, recruitment strategies, analytic techniques and, finally, ethical issues associated with my study.  3.1 Research questions  As discussed at the end of the literature review, the two research questions (RQs) that guided this study were as follows: • RQ1: What are the different ways that weight stigma in health care can be conceptualized?  • RQ2: What strategies can be employed to reduce weight stigma among health care providers in BC?  These questions were developed through an examination of gaps in the literature and were investigated through the lens of a case study of BalancedView. Through research question one, I aimed to enhance understandings of weight stigma in health care. In particular, I explored perspectives among participants in this case study on what weight stigma is and what it looks like in health care. For example, I was interested in how health care providers involved in developing or taking the BalancedView course viewed weight stigma and how other stakeholders involved in curriculum development conceptualized weight stigma in health care (e.g. patients 61  and experts interviewed as part of course development or directly involved in developing the course). I also looked at what the areas of agreement and disagreement were.  Through research question two, I reflected on the range of ways considered within the case study to address weight stigma within health care in BC and on the effectiveness of the various strategies used within BalancedView to reduce weight stigma among participants. To guide my inquiry around question two, I explored what strategies were identified as possibilities to reduce weight stigma. I also looked at what strategies were considered politically acceptable and feasible. Based on the implementation of BalancedView, I additionally investigated what strategies were most influential in reducing weight stigma among participating health care providers. Next, I describe my research approach, which was based on a case study and drew on a pragmatic paradigm.  3.2 A case study research approach  My research questions were explored within the context of a case study on weight stigma reduction among health care providers in BC and the case of BalancedView. I aligned my case study with the methodological traditions of a ‘pragmatic approach’ (or ‘paradigm’) (Morgan, 2007). According to Merriam (1998), while case studies are common, there is less clarity or consensus on what a case study is or how it should be done. Yin (2003, 2014) similarly notes that case studies are regularly used, though often criticized. He further argues that many of the criticisms of case study research stem from a lack of clarity about what a case study is. In this study, a case study was defined as: 62  . . . a strategy for doing research which involves an empirical investigation of a particular contemporary phenomenon within its real life context using multiple sources of evidence. (Robson, 2002, p 178) As the above quote suggests, case studies are an approach to research – rather than a data collection method – that are characterized by investigation of real life occurrences, within a particular and specified context (Baxter & Jack, 2008). A case study design is considered for selection when you cannot manipulate or control the behaviour of those involved in the study or when contextual conditions are believed to be pertinent to the topic at hand (Pearson, Albon & Hubball, 2015; Yin, 2014). In my proposed study, I was one of many stakeholders37 involved in developing the BalancedView course to address weight stigma and examining what works to mitigate it in health care, hence I had only a limited degree of control over the course and research design. As Harland (2014) states, case studies are beneficial when learning from specific cases may help better understandings of complex, real life phenomena. Weight stigma is a complex, real life issue in health care. Looking at applied attempts to reduce it in a real world setting was an invaluable opportunity for advancing knowledge.  Further, a case study approach was relevant as my research was specific to weight stigma among health care professionals in BC. I did not presume to generalize my findings to all health care professionals. Rather, I aimed to explore the issue of weight stigma in the health sector                                                 37 Three committees were involved in developing BalancedView: A steering committee of experts, an advisory committee of health care providers and, to a lesser extent, the Promoting Healthy Weights Working Group. These committees are described later in this chapter.  63  within the BC context and, more specifically, within the boundaries of the development and implementation of BalancedView. However, as Harland (2014) points out, while replicability – something that is often thought to be important in weight stigma reduction intervention research – cannot occur through case studies, lessons can nonetheless be gleaned that are relevant to professionals and academics in the same area. My hope is that the lessons discussed in this dissertation, while bound up in the context of BalancedView, are useful to other professionals or researchers wanting to undertake similar initiatives to reduce weight stigma in health care.  In addition to the emphasis within case studies on phenomena in real life contexts, another central feature of case studies is the use of a variety of data sources or multiple methods (Baxter & Jack, 2008; Harland, 2014). Use of multiple data sources and methods, as Baxter and Jack (2008) argue, “[e]nsures that the issue is not explored through one lens, but rather a variety of lenses which allows for multiple facets of the phenomenon to be revealed and understood” (p 544). This aligns well with the mixed methods I used in my study.  I primarily drew on Yin’s (2014) understanding of case study research. While others, such as Merriam (1998) and Stake (1995) have also discussed the case study methodology, I gravitated towards Yin’s (2003, 2014) work as he argues that the philosophical beliefs between quantitative and qualitative research are reconcilable and there is a “strong and essential common ground between the two” (Yin, 2003, p 15). This fits with my mixed methods, pragmatic approach to this study. Yin also provides helpful insight into how to design and practically conduct a case study, as well as criteria through which to evaluate case studies.  The type of case study I used to explore the problem of weight stigma in health care was an exploratory-explanatory study, merging two of the forms of case studies described by Yin 64  (2003). Exploratory case studies are helpful when the phenomenon under investigation is relatively underexplored and about which not enough is known to develop formal hypotheses. This type of case study is also useful for exploring ‘what’ questions and can be used as a precursor to more formal explanatory studies. In an exploratory design, the researcher need not have pre-determined theoretical propositions before beginning research (Streb, 2010). In my case, the exploratory part of the study was around problem definition and conceptualizing weight stigma. Explanatory case studies, by contrast, are about theory testing and building (Yin, 2014). The explanatory part of my study aimed to explore how to reduce weight stigma among health care professionals, based on the case of BalancedView.  An important element of case study design is the idea of case boundaries (Yin, 2014). The researcher must determine what fits within the scope of the case study and what does not. My case study focused exclusively on data collected through the development, pilot testing and implementation of BalancedView. As such, I used a single case study design, insofar as I examined the issue of weight stigma in health care through the case of one intervention project rather than multiple projects. Within my single case study design, I drew on multiple units of analysis (Yin, 2014).38 The units of analysis included health professionals participating in the course, stakeholders involved in developing the course, as well as the course itself and all                                                 38 Yin (2014) outlines that regardless of what type of case study is selected, the case study can be either single (i.e. focused on one case) or multiple (i.e. focused on more than one case) and that the researcher should decide on this during the design phase. He articulates that single cases are most helpful for understanding the peculiarities of a specific case, while multiple cases are helpful for building theory. Within either subtype, the approach can be holistic or embedded. Holistic case studies refer to when the case(s) at hand are studied from a general perspective, as a whole. Embedded case studies are when the there are multiple units of analysis (e.g. multiple people of interest who are studied in their own right within the broader case). The benefit of having multiple units of analysis, particularly within a single case study, is that it allows room for theory generation across the units despite the focus on a single case. 65  associated materials. My own reflexivity also informed the analysis, which I discuss later in this chapter. In order to maximize the reliability of case study analyses and findings, researchers should specify early on what criteria will be used to determine the trustworthiness of the interpretations (Yin, 2003). The below table summarizes the criteria I used for interpreting findings.  Table 1 Case study criteria Criterion Description Member checking Were interpretations checked with members? What alternate interpretations or points of contention existed, if any?  Reflexivity Have I considered and been explicit about how my personal experiences, thoughts and biases influenced my interpretations? With this case study approach in mind, I next discuss my research paradigm.  3.3 Informed by a pragmatic research approach  I used a pragmatic research approach – or pragmatic paradigm – for this case study. The word ‘paradigm’ is used often in the social science literature, although there is not necessarily agreement on how to conceptualize it. Morgan (2007) outlines three main trends in how paradigms are considered in the research literature. One trend is to broadly consider a paradigm as a ‘worldview’, another to consider a paradigm as a philosophy of knowledge related to one’s 66  epistemological and ontological stance.39 A third is to consider a paradigm as a set of “shared beliefs within a community of researchers who share a consensus about which questions are most meaningful and which procedures are most appropriate for answering those questions” (Morgan, 2007, p 53). Morgan (2007) argues that the first of these definitions is overly generalist and that the second is divisive in that if we conceptualize a paradigm as an epistemological stance that can have little cross over with other ways of knowing, we create disciplinary silos and ignore potentially valuable ways of knowing. He advocates that instead we adopt the third conceptualization of a paradigm and connect that with ‘pragmatism’ (Hookway, 2013). I located my work within Morgan’s notion of a paradigm as a set of shared beliefs about what questions are important to study and how to go about exploring those questions. I did this because I did not want to limit my work to one ontological or epistemological position. I was interested in treating all stakeholders’ perspectives and various knowledge sources as valid in order to further our shared goal of better understanding and reducing weight stigma. In the words of Russell and Cameron (2016) “. . . different disciplines offer useful insights that become even more powerful when brought together, fat pedagogy is, and must be an interdisciplinary, multidisciplinary, and transdisciplinary endeavour” (p 254). I also connected my work to Morgan’s notion of a pragmatic paradigm. Morgan conceptualizes a pragmatic paradigm of research as being focused on action – much like this study – rather than epistemology. Key foci of this research paradigm, according to Morgan (2007) are action and the beliefs behind these actions. The emphasis is on both shared meaning                                                 39 The notion of a paradigm as an epistemological stance was promoted largely through the qualitative revolution beginning in the late 1970s and continued into the early 2000s, with researchers like Guba and Lincoln (1994, 2005) at the helm. In that era, we saw a (re)emergence of qualitative research and a recognition of the validity of such research within social science. 67  and joint action. This approach was beneficial to me as it is inherently collaborative. It focuses on mutual understanding where possible. Despite the focus on mutual understanding and action, a pragmatic paradigm of research does not necessarily abandon issues of epistemology. Rather, it is consistent with a focus on epistemology when practical or helpful. This means that the focus is not on the philosophy of knowledge in its own right. This is better left to philosophers, Morgan (2007) argues. Instead, in a pragmatic framework researchers should consider how beliefs about different ways of knowing constrain or enable us in understanding and solving problems. Thus, within my analysis, I considered how the various epistemological stances of stakeholders involved in this project influenced the discourse around weight stigma. In the spirit of pragmatism, however, I was not interested in prioritizing one way of thinking over another, as the study of a complex social issue like weight stigma is best understood through multiple perspectives.  3.4 The case of BalancedView In response to the problem of weight stigma in health care and the need to develop knowledge on how to reduce this bias, the Provincial Health Services Authority (PHSA) funded, developed, pilot tested, implemented and evaluated an online anti-weight stigma course for health care providers in BC, in partnership with an array of stakeholders (described below). The course, which is still publicly available, is entitled BalancedView (BV).  BalancedView is a multi-module course that aims to reduce weight stigmatizing attitudes among participants, enhance knowledge about weight stigma and build competency to address weight stigma in clinical practice. The target population are health care professionals in BC who self-identify as belonging to one of the following groups: 1) medical professionals (e.g. 68  paediatricians, general practitioners); 2) mental health professionals (e.g. psychiatrists, psychologists, social workers, counsellors, child and youth care workers, mental health clinicians); or 3) allied health care professionals (e.g. occupational therapists, recreation therapists, dietitians, nurses, other).  The development of BalancedView and its implementation and evaluation was funded by the Provincial Health Services Authority’s (PHSA) Population and Public Health Program. The contract to undertake this work was housed with the Health Literacy team at BC Mental Health and Substance Use Service (BCMHSUS), an agency of the PHSA.40 After extensive work developing the course, it was then piloted and launched. Each phase is then described in detail in its own right, beginning with the development.  3.4.1 Development description  The BalancedView project began in fiscal year 2012/2013 with PHSA funding to develop, pilot test and implement a three-year educational resource for health care providers in BC. Given the limited amount of information about how to reduce weight bias in health care, the team decided to utilize a participatory approach to developing the resource, such that the wisdom of an interdisciplinary team could be incorporated. Three committees were actively involved in developing the project: 1) a steering committee of professionals and researchers with expertise in weight stigma and related topics (e.g. eating disorders prevention) (membership fluctuated from 11 to 17 people); 2) an advisory committee of diverse health care providers across the province                                                 40 BCMHSUS is no longer the project lead for BalancedView. Due to larger organizational shifts within the PHSA, BalancedView is now the purview of the Health Literacy team at BC Children’s Hospital. The Health Literacy team formerly existed within BCMHSUS, but as of this writing the team falls within BC Children’s Hospital, which is also an agency of the PHSA.  69  (membership fluctuated from 10 to 13 people); and 3) a pre-existing BC committee with a mission to prevent weight stigma and eating disorders, entitled the Promoting Healthy Weights Working Group (PHWWG), with an open membership and fluctuating numbers that met monthly, that I previously sat on.41 An estimated 60 hours was spent consulting with these groups over the course of development. Meeting schedules varied over the course of the project, sometimes occurring with the committees monthly, especially in the earlier days, and then moving to every few months as needed.  A single evaluator consultant was initially hired to evaluate the project, a role which I was later invited to share given my research interest in the topic and intent to make this my PhD project. The evaluator and I worked with the team to establish desired outcomes from the project and a plan for measuring success.  It was decided through a collaborative process at the committee meetings that, building on the funding proposal from BCMHSUS, the overarching activities of the project would be:  1) Perform a needs assessment (literature review, environmental scan and gap analysis) to understand the needs of health professionals, patients and families in British Columbia (BC) with regard to weight bias and stigma. 2) Develop an evidence-informed resource to address weight bias and stigma among health professionals in BC. 3) Disseminate the weight bias and stigma resource to health professionals in BC.                                                 41 The PHWWG was less involved in the development of BalancedView and decision-making about the course than the other committees.  70  4) Evaluate the weight bias and stigma resource. (PHSA, 2013, January)42  Corresponding to these objectives, the committees worked with the evaluation and project management team, including myself, to develop a set of desired outcomes as set out within an outcomes measurement framework. These outcomes were: “Health care professionals have increased knowledge, enhanced attitudes and competencies regarding weight bias and stigma” and “[h]ealth professionals implement changes in their care of patients” (PHSA, 2015, April). While discussion took place about the merits of online learning versus in-person forums, due to budgetary constraints and a desire for sustainability, a decision was made early on by the project management team to utilize an online format for the course. Two contractors – with little professional or academic involvement on the subject of weight stigma, but with extensive research backgrounds – were then hired to conduct a ‘scoping review’ that explored why weight stigma exists and what could be done to address it in health care. This scoping review included a systematic literature review and 22 interviews (three with patients and the other 19 with individuals with subject matter expertise). These contractors worked with the three committees and project managers to collaboratively develop a final report from the scoping review. The scoping review43 acknowledged that weight stigma was a complex problem that would likely be best solved through a multi-method approach to stigma reduction. It also articulated a relationship between the medicalization of weight and weight stigma.                                                   42 Quote as per the Project Charter.  43 The search strategy involved MEDLINE, CINAHL, PubMed, SocINDEX, PsycINFO, Social Services Abstracts and Social Work Abstracts. The search was based around themes of health care professionals, health care delivery, overweight, obesity, stigma and bias, and reducing stigma and bias. For more information see PHSA (2013a). 71  Based on this report, another contractor was hired to work with the committees and project managers to develop a Word document with the desired content that would ultimately be translated into the online course. The contractor who was hired to do this was an academic who was a proponent of Health At Every Size principles and had a background in dietetics and education. Over several months this content contractor worked with the committees, project managers and myself to draft the content, which included information on medicalization and a strong emphasis on a community of practice that aimed to facilitate user engagement beyond the course ending. After this, it was collaboratively determined that the content was ready to be sent to the IT company to translate it into an online format. Months later, in December 2013, a daylong meeting with both the steering and advisory committee was held to review the preliminary translation of the content to the online format. At this meeting, the IT company presented a condensed version of the content document to the committees structured into a format that made sense visually. The modules included in the content map they presented were as follows:  • Module 1: Understanding weight stigma (a basic overview of weight stigma) • Module 2: Patient voices (video recordings of patient stories) • Module 3: Weight science, alternative paradigms and ethics (information contesting common ‘myths’ about overweight and obesity, like overweight and obesity leading to disease and early death) • Module 4: Professional voices (medical professionals on film sharing their stories of change) • Module 5: Applying alternative paradigms (e.g. HAES) (how to apply a health-centred, de-medicalized approach to working with heavier patients) • Module 6: Realities of practice 72  • Module 7: Community of practice (a discussion forum to encourage engagement and ongoing learning)  The intention at this point was to solicit input into the look and feel of the proposed online format of the tool from the committees. However, feedback from some committee members focused more on content, despite the seeming consensus on content that had previously been achieved. At the December 2013 meeting, several committee members expressed discomfort at the sections that contested the common belief that overweight and obesity led to health problems. Given this disagreement, over the next several months, the content around medicalization was revisited. Ultimately, it was decided that the content focusing on medicalization would take a more neutral perspective and would present both sides of the argument on medicalization, including its possible benefits and consequences.  After the revision process, the course was finalized at five modules and the contentious content on medicalization was de-emphasized to a large extent.44 The community of practice was also excluded, as the resource was lengthy. These five modules take about 2.5 hours to work through. The course itself is interactive and multi-medium, using text and audio-visual. There is an audio-visual host (e.g. a hired actor) who guides participants through the training. There are learning activities and assessments at the end of each module (e.g. quizzes, shared scripted reflections) and videos (including an interactive video where participants pause to make comments).  During the development phase of BalancedView, the other evaluator and myself worked with the committees to develop pre-course and post-course questionnaires for users to take                                                 44 A reflexive analysis that addresses my feelings about this occurs later.  73  immediately before and after the course. Once both the questionnaires and modules had consensus, we moved to the pilot test phase.  3.4.2 Pilot test description Early in the development of BalancedView, a plan was formulated to pilot the course with health care providers prior to officially launching it. Initially it was hoped that the pilot would be an opportunity to test the strategies employed in BalancedView in terms of their effectiveness in reducing weight bias and to make revisions accordingly. However, due to the significant financial and time resources that went into the creation of BalancedView, including filming hired actors at union rates and producing videos, it was determined that to make substantive changes to the resource based on pilot feedback was not economically or technically feasible. It was decided instead that the aim of the pilot would be to work out any glitches in the online platform and to allow for feedback on the questionnaires and any text-based parts of the course that could be simple and cost effective to change.  The five BalancedView modules and the questionnaires were pilot tested in late 2014 after the above-mentioned extensive process of development and revisions by the committees. The pilot recruitment strategy was collaboratively decided through discussions in various committee meetings and included recruiting five people from each health authority (for a total of 25 pilot participants), with the aim of having diverse professional representation. Monetary incentives were offered for completion of the course and questionnaires. Twenty three of the 25 initial recruits registered for the pilot test. Due to many technical problems, few people completed it and another round of pilot recruitment occurred, with an additional seven people 74  recruited. Of the two rounds, only eight completed the course, largely due to technical issues with the platform. These technical issues were mostly resolved post pilot.  Although based on a small number of participants, the pilot test results suggested the resource was effective in decreasing weight-biased attitudes and possibly helpful in changing behaviours in the future, although no participants had yet implemented any changes to their practice immediately after the pilot. Participants found the course content useful and important and had few suggestions for how to improve it. The minor suggestions for content change included having more patient stories and more information on the medicalization of weight. The main area for improvement suggested by participants related to technological issues, which were significant. As a result of the pilot, additional months were spent ironing out technical glitches. Participants also suggested shortening the questionnaires.45   3.4.3 Launch description  On March 31, 2015, the last day of the official three-year funding package from PHSA, the course was publicly launched. As of September 2016, at which point I stopped actively collecting data, BalancedView had been widely disseminated and implemented. A total of 925 registrants had signed up for BalancedView by September 6, 2016 (including pilot participants). Of the larger group, some were excluded from the case study, for reasons including: location outside of BC or Canada (thus out of scope); incomplete registration profiles leading to inability to assess for study eligibility; registrants with administrative roles in the BalancedView development; or registrants who were students and not yet health care providers. After excluding                                                 45 Several questions were removed from the questionnaires after the pilot, discussed shortly. See Appendix B, part four to view questions that were removed.  75  these participants, as well as the pilot participants, 654 eligible participants enrolled in BalancedView. A total of 249 participants from this larger cohort completed the course (38.07% completion rate) and were included in the present analysis. From this group, 74 participants were given monetary incentives to also participate in three and six-month follow-up.46 Fifty-six people completed the three-month follow-up and 46 completed the six-month follow-up.  3.4.4 The course content  The final course, still accessible at the time of this writing at balancedviewbc.ca (BC Mental Health and Substance Use Services, 2015), involves several main subsections, as follows:  • Module 1: Understanding Weight Stigma o Implicit Association Test (IAT) o Introduction to weight stigma  • Module 2: Patient Voices o Videos of real patients sharing stories  • Module 3: Introduction to Health-Centred Approaches  o Medicalization pros and cons Prezi o Video with medicalization debate  o Introduction to health-centred approaches  • Module 4: Professional Voices  o BC health care providers sharing stories of change  o Rudd Center video  • Module 5: Applying Health-Centred Strategies in Practice  o Video scenarios with one interactive video to practice applying health-centred strategies                                                  46 These 74 participants were provided with $50 for completing the pre-course and post-course questionnaires, $25 for participating in the three-month follow-up and another $25 for participating in the six-month follow-up.  76  Module one seeks to raise awareness of weight stigma and its harms, including critical reflection of one’s own beliefs (and perhaps thus cultivate self-awareness and cognitive dissonance). The Implicit Association Test (IAT), the first activity in module one, is intended as a tool for participants to reflect on their own, perhaps implicit, weight bias. Module two strives to evoke an empathic response among learners through viewing and listening to the stories of several real patients in BC talking honestly about their experiences of weight stigma in health care. Module three aims to introduce the concept of medicalization and challenge learners to think critically about weight-centred approaches to health, the possible harms of medicalizing weight and the benefits of adopting a health-centred approach. In doing so, this module presents both the harms and possible benefits of medicalizing weight, such that viewers are presented with more than one perspective on the topic and encouraged to think critically. This module also aims to challenge common assumptions about weight, particularly the notion that weight is easily controllable, as per attribution theory. Module four uses videos of professionals sharing their stories of shifting their own beliefs and practices around weight and is aligned with social consensus theory. Module five has interactive activities designed to help health care providers practice being unbiased with fabricated video scenarios intended to resemble actual practice scenarios. Each module has learning activities, including quizzes or scripted reflections. Various videos were filmed for the project in addition to the patient and professional voices videos to maximize learning. Videos are described in a table in the first findings chapter (see p 163). With this contextual information about the case in mind, I now discuss my data collection methods. 77  3.5 Data collection methods In keeping with a case study and pragmatic approach, my study drew on mixed methods. Mixed methods involve using more than one method, often drawing from both the qualitative and quantitative tradition and are helpful in order to provide more than one way of looking at a particular issue (Stockman, 2015). In the weight stigma intervention field, studies with a pre- and post-test methodology and validated quantitative measures are advocated (Daníelsdóttir et al., 2010). However, as Link et al. (2004) discuss, qualitative research can help us deepen understandings of the intricacies of stigma. My study thus brings together qualitative and quantitative methods, with an emphasis on the former. My data collection tools and their relationship to my research questions are highlighted in the below tables and subsequent image.    78  Table 2 Research question one and associated data collection methods RQ1: What are the different ways that weight stigma in health care can be conceptualized? Method Description   Participant observation  Participant observation occurred during the development phase of BalancedView of steering and advisory committee and external reference group meetings and interactions. This helped to examine how health care providers and experts involved in developing BalancedView conceptualized weight stigma. Document analysis Document analysis was undertaken of documents associated with BalancedView (e.g. course text, meeting minutes, content document, scoping review, qualitative questionnaire data, scripted reflections on BalancedView) to further assess how various stakeholders and course participants conceptualized weight stigma. Research journal  A research journal was kept in which I took field notes following meetings with the project team or committees. My research journal incorporated a reflexive component. I documented how I saw weight stigma being spoken about and how my feelings influenced my interpretations.  Group interview  One group interview occurred with two members of the committees who were actively involved in course development. Data from the interview informed my analysis of how the committees conceptualized weight stigma.   Focus group  A focus group with steering and advisory committee members helped explore their perceptions of weight stigma. Semi-structured interviews with course participants  Interviews with course participants were conducted through which their perspectives of weight stigma were explored (pilot n = 3; post pilot n = 10).    79  Table 3 Research question two and associated data collection methods RQ 2 What strategies can be employed to reduce weight stigma among health care providers? Method Description   Document analysis   A document analysis was conducted of documents including the scoping review and committee meeting minutes to understand strategies considered to reduce weight stigma during development of BalancedView. After the implementation of BalancedView, the document analysis also included the scripted reflections in BalancedView and qualitative questionnaire responses in order to explore course participants’ experiences with the different stigma reduction strategies. Participant observation  Participant observation was conducted at the various committee meetings. I took reflexive field notes in my research journal to document and analyze strategies considered during the development of the course to address weight stigma. Questionnaires  Online questionnaires were administered to course participants right before and right after BalancedView (249 participants completed both pre-course and post-course questionnaires). Online questionnaires were also administered at three-month (n = 56) and six-month (n = 46) follow-up. The questionnaires helped explore the effectiveness of BalancedView and understand participants’ perceptions of most useful aspects of course. Interviews  Interviews with participants who took the course were conducted to understand their experiences with the different stigma reduction strategies in the course (pilot n = 3; post pilot n = 10). Focus group and group interview  A focus group and group interview with committee members were conducted regarding stakeholder perceptions of what helps to reduce weight stigma among health care providers.   80  Figure 1 Image of methods used in each phase of BalancedView  Each method described in the above table and featured in the image is elaborated upon below.  Development of BalancedViewPilot test of BalancedViewImplementation of BalancedView with 249 participants Follow-upParticipant observation at committee meetings pertaining to developing BalancedViewResearch journalDocument analysisGroup interview with committee membersFocus group with committee membersQuestionnaires before and after pilot (n = 8)Interviews with pilot participants (n = 3)Pre-course questionnaires with 654 participants (only 249 finished course)Post-course questionnaires (N = 249)Three-month follow-up questionnaires (n = 56)Six-month follow-up questionnaires (n = 46)Scripted reflections (N = 249)81  3.5.1 Participant observation, reflexivity and research journal  An integral aspect of my research was participant observation of those involved in the development of BalancedView. Participant observation is a method that involves observing behaviour and interactions, listening to conversations and, oftentimes, asking questions. It can be undertaken covertly, overtly, in public settings or in environments that are private and more challenging to access (Bryman, 2008). Participant observation helped inform my findings of how weight stigma was conceptualized during the development phase of BalancedView and the range of possible ways weight stigma could be addressed in health care in BC that were discussed by the committees.  I regularly attended and participated in meetings with the project management team and the often-held advisory and steering committee meetings and undertook participant observation. The frequency of committee meetings varied depending on need, ranging from monthly to every few months. In total, the project management team estimated 60 hours of committee meetings over the course of the project. Five of the meetings were recorded with permission and I had access to the audio files to review at a later date to supplement my analysis. It was at these meetings that integral decisions were made about the project47 and important discussions took place about weight stigma. I documented my observations in a research journal and kept notes of what happened at meetings, what I learned and what my thoughts and emotions were and why. My research journal was also infused with a reflexive component (discussed more in the analysis section of this chapter) where I tracked thoughts, feelings and interpretations. Within this I strove                                                 47 Decision-making was a collaborative process with committee members, however, final decision-making authority rested with the PHSA, who had representation at all committee meetings.  82  to consider the ways in which my opinions were influenced by my past experiences and social location. As I show in my findings later, weight stigma appears to be intertwined with emotions. Given the emotional nature of the subject matter as well as my need to capture my own feelings throughout the process, I found myself at times using elements of arts-based research in my research journal, despite not intending this at the outset. Arts-based research draws on the medium of art, such as poems, visual art, theatre or otherwise to explore information that more traditional qualitative or quantitative methods cannot (Leavy, 2009). In my case, writing poems in my research journal helped me to capture my emotionally laden experiences in a way that my other methods did not.  3.5.2 Documents as sources of data  Bowen (2009) defines document analysis as a systematic procedure for collecting, reviewing and interpreting data. It is particularly useful in combination with other research methods as a form of triangulation (Bowen, 2009). I included all types of documents accessible to me48 that were part of the development and implementation of BalancedView, for example: meeting minutes; background project documents; research journal entries; online scripted reflections by participants during BalancedView; and the qualitative sections from questionnaires (described as a method in their own right shortly). The project team developed the scripted reflection questions, with input from the steering and advisory committees.49 The reflections asked BalancedView course participants to reflect on aspects of the course, such as how they felt learning their Implicit Association Test scores, their perceptions of the ways that the                                                 48 Documents were included if they fit within the scope of the case boundaries and related in some way to the research questions.  49 I provided input on the scripted reflections but did not take the lead.  83  medicalization of obesity has impacted their patients, either negatively or positively, and to comment on the videos. For a complete snapshot of the scripted reflection questions see Appendix A. A huge amount of qualitative data was collected for document analysis: 37 official meeting minutes from the advisory and steering committees and working group meetings; 80 project background documents (including: multiple iterations of the evaluation outcomes measurement framework, terms of reference for the committees, scripts for audio scenes in the course and the scoping review); 58 journal entries; over 90,000 words of online scripted reflections from participants who took the course; over 50,000 words of qualitative text from participants responses to the pre-course questionnaire, over 60,000 words in the post-course questionnaire; and over 7,000 words of qualitative text from follow-up questionnaires.  3.5.3 Questionnaires  The question of ‘what works to reduce weight stigma?’ has primarily been taken up by social psychologists in experimental studies that use quantitative methods and questionnaires with attitudinal measures (Daníelsdóttir et al., 2010). The current study used pre-course and post-course questionnaires to explore the effects of the course on participants. The pre-course questionnaires were taken by participating health care providers right before starting the course and the post-course questionnaires were completed immediately after the course. Using pre- and post-test measurement is important given the critique by Daníelsdóttir et al. (2010) that studies failing to use some form of pre-test measurement have limited interpretability. The questionnaires were developed through a review of measures in the weight stigma literature and through collaborative conversations at committee meetings. The types of questions asked in the 84  questionnaires were based around how the BalancedView team conceptualized weight stigma, which is articulated in the first findings subsection. This was important as “. . . the validity and reliability of a measure cannot be assessed unless the concept being measured is defined in a clear unambiguous fashion” (Charles & White, 2008, p 78).  The questionnaires were tested during the pilot phase with eight participants. Pilot participants had no substantive comments on the questionnaires other than the rather lengthy nature. Questionnaires were shortened following the pilot.50 The final versions of the questionnaires were administered online to course participants immediately before and after taking the course. In total, 654 eligible participants took the pre-course questionnaire, however, only 249 completed the course and the post-course questionnaire. These 249 participants are the main focus of the analysis. This study did not use a matched sample. Three and six months after the course, follow-up questionnaires were sent to a smaller sample (n = 74) recruited by the project managers for long-term follow-up. Among this group, 56 completed the three-month follow-up and 46 completed the six-month follow-up.51 The full questionnaires are available in the Appendix B. Next, I discuss the content of the questionnaires, beginning with demographics. Demographic questions Prior to taking the course, participants were asked for demographic information including: whether they work for a health authority; which health authority they work for (if applicable); organization employed by (if not employed by health authority); location;                                                 50 As a result of the feedback we removed one 9-point scale we had created with respect to medicalized practices in health care (see Appendix B, part four). We also removed the pilot-specific questions asking participants about their experiences with the questionnaires, learning assessments and scripted reflections. 51 Selection criteria for participation in follow-up are discussed shortly.  85  profession; role; length of employment as a health professional; age groups worked with; whether they work regularly (i.e. at last 50% of work time) in any of the following areas – eating disorders, obesity, cardiac, diabetes, weight loss or bariatric surgery;52 size of population area worked in (e.g. small population centre [population 1,000 to 29,999] to large population centre [100,000+]); gender; and age.  Attitudinal questionnaire measures  Implicit and explicit attitudes were measured through the questionnaires. In the post-course questionnaire, participants were also asked to rate their agreement with the following statement on a five-point agreement scale: “Participating in BalancedView has contributed to decreasing my weight-biased attitudes” (response options from 1 = strongly disagree to 5 = strongly agree). The team also drew on measures from the literature to measure explicit attitudes, including the short form of the Fat Phobia Scale and the Attitudes about Treating Obese Patients scale.  The 14-item Fat Phobia Scale (Bacon, Scheltema & Robinson, 2001) is a measure in the literature that has been found to have good to excellent reliability with a Cronbach’s Alpa of .87 from a 1984 to 1991 sample and a Cronbach’s of .91 from a 1999 sample. The Fat Phobia Scale has 14 bipolar adjectives (e.g. lazy / industrious; attractive / unattractive), each along a five-point line. Participants indicate the point along that five-point line that best describes their feelings and beliefs about fat people. Items following a positive to negative trait code are scored as 1 to 5; items following a negative to positive trait code are reverse coded. Reportedly, scores of 2.5 or                                                 52 The questionnaires did not explore whether participants had personally experienced any of these issues. However, the post-course questionnaire did include one question on self-reported height and weight and one question on whether they had ever been teased for their weight.  86  higher are considered weight-biased (Bacon at al., 2001). We used this measure in the questionnaires administered right before and right after BalancedView and at three- and six-month follow-up. In addition to drawing on this 14-item measure, based on the specific weight-biased attitudes we wanted to address through our intervention, we added two additional questions (intelligent / unintelligent; healthy / unhealthy) to the scale. Results are presented for both the 14- and 16-item measures used. Although we adapted the scale, the Cronbach’s Alpha for the 16-item was excellent in the pre-course measurement (.90) and good at the post-course measurement (.88). The Cronbach’s alpha for and the 14-item also showed good reliability at the pre-course measurement point (.89) and at the post-course measurement point (.89).  We also adapted and used five questions from the Attitudes about Treating Obese Patients scale for use in the pre-course questionnaire, post-course questionnaire and the three- and six-month follow-up questionnaires. This is a more recently developed 18-item tool to explore views health care practitioners have about obese patients (Puhl, Latner, King & Luedicke, 2014).53 In this scale, each item is framed as a statement about patients with obesity and participants are asked to rate the extent to which they agree with the statement on a 5-point likert scale (from 1 = strongly disagree to 5 = strongly agree) with statements about obese patients (e.g. “I feel that obese patients are often non-compliant with treatment recommendations”). The benefit of this tool is it is health care specific. However, given our already lengthy questionnaire, we utilized only a portion of this scale, as there were concerns that if the pre-course questionnaire was too long then attrition might increase. Five questions were                                                 53 There are two subscales, one on negative attitudes about patients with obesity (12 items, Cronbach’s Alpha .89) and the other on frustrations with treating patients with obesity (6 items, Cronbach’s Alpha .80) (Puhl et al., 2014).   87  used from this scale. The five questions were slightly modified from the original measure in that the term ‘obese patients’ was rephrased to ‘patients with obesity’, at the request of some committee members who preferred to use person first language. The five items used were as follows: 1) I feel that patients with obesity are often non-compliant with treatment recommendations; 2) I feel that patients with obesity lack motivation to make lifestyle changes; 3) Treating a patient with obesity is more frustrating that treating a non-obese patient; 4) It is difficult to feel empathy for a patient with obesity; and 5) I would rather treat a non-obese patient than a patient with obesity. Despite the modifications described above and only using five of the items as a measure of participants’ weight bias, we found that these five items had a Cronbach’s alpha of .78 at the pre-course questionnaire measurement point, suggesting an acceptable level of internal consistency as a measure of reliability (Tavakol & Dennick, 2011). From the post-course questionnaires the Cronbach’s Alpha was .86. In addition to these five items, we also asked participants two questions about perceptions of their colleagues’ weight bias, also drawn from the same study by Puhl et al. (2014). Puhl et al. (2014) asked participants to rate their peers’ weight bias alongside the Attitudes about Treating Obese Patients scale. The two questions we used, which were adapted from Puhl et al. (2014), were: 1) Other health providers in my field often have negative stereotypes toward patients with obesity;54 and 2) I have heard/witnessed other professionals in my field make negative comments about patients with obesity. These questions were seen as relevant to get a sense of the difference between one’s own self-reported bias versus perceptions of colleagues’ bias. Of note, these                                                 54 In the Puhl et al. (2014) study this question was specific to eating disorder professionals and read: “Other practitioners who treat eating disorders often have negative stereotypes about obese patients” (p 4). Both questions were modified to be person first (e.g. ‘patient with obesity’ not ‘obese patient’).  88  questions were only asked in the pre-course questionnaires since the responses were unlikely to change after the course.  A measure of implicit attitudes was also used in this study in at baseline, in the post-course questionnaire and in the three- and six-month follow-up questionnaires. In response to the criticism that explicit measures of weight stigma are subject to response bias (Daníelsdóttir et al., 2010), Schwartz, Chambliss, Brownell, Blair and Billington (2003) and O’Brien et al. (2010) utilized the Implicit Association Test (IAT) – a form of pattern matching – as a measure of implicit weight-biased attitudes. The IAT is a timed word association test (words and images have to be sorted) that has a paper and online version and has been tested extensively (Greenwald, Nosek & Sriram, 2006). After taking the test participants receive a bias rating of one of the following: strong automatic preference for thin vs. fat; moderate automatic preference for thin vs. fat; slight automatic preference for thin vs. fat; little to no automatic preference for thin vs. fat; slight automatic preference for fat vs. thin; moderate automatic preference for fat vs. thin; or strong automatic preference for fat vs. thin. If participants have too many errors or are slow in their response time a score may not be able to be computed.  In the context of an online course we wanted to utilize the online version of this test. An American company provides access to the online IAT for individual use at no cost and reports scores back to individuals. However, to access raw group data we were informed it would cost us an amount that was outside of the BalancedView budget allocation. Utilizing an American company also raised privacy issues: no personal information can be collected or stored in the United States as per PHSA privacy legislation. To get around this we decided that participants would link to the online IAT (on the Project Implicit® website) in the first module and again in 89  the post-course questionnaire and that based on their company-reported individual scores, they would self-report their result back onto the BalancedView platform. Response options and accordant bias scores used in this study were: strong automatic preference for thin vs. fat (bias score = 3); moderate automatic preference for thin vs. fat (bias score = 2); slight automatic preference for thin vs. fat (bias score = 1); little to no automatic preference for thin vs. fat (bias score = 0); slight automatic preference for fat vs. thin (bias score = -1); moderate automatic preference for fat vs. thin (bias score = -2); or strong automatic preference for fat vs. thin (bias score = -3). Of note, participants completed the baseline IAT as the first activity in module one rather than in the pre-course questionnaire. This was chosen in order to reduce the time burden of the pre-course questionnaire and decrease the likelihood of attrition.  Other questionnaire measures  In addition to attitudes, the project team also wanted to assess competency in addressing weight stigma in clinical practice. Competency, a construct which was defined with the committees as including knowledge, skills, behaviours and reflexivity, was measured via self-report questions and via questions on how providers interact with patients of various weights. The inclusion of questions around competency helped avoid the common pitfall in weight stigma research in which attitudes are presumed to lead to behaviours, though not measured as such. The questionnaires also ask questions about the course content and delivery method in order to explore which aspects of the course were most influential and why. For instance, participants rated components of the course with a drop-down list to provide information on what aspects were most influential to their learning.  90  In addition to these measures, the post-course questionnaire also contained some opportunities for the participants to provide qualitative responses. For instance, regarding the statement “After taking this course, I am able to identify ways to avoid weight stigma in my practice”, participants responded based on an agreement scale and were asked to explain their response. Another question pertained to whether they had yet made changes to their practice (“Have you yet implemented any changes in your practice as a result of this course. If so, please describe”). Yet another question related to what they intended to do differently in their practice (“Please describe what you will do differently in your practice, if anything, after taking this course”). Questions were also asked about the course itself, including: “What did you find most useful about this course?”; “Please describe any factor or situation that impeded your learning in this course”; and “How could BalancedView improved?”  3.5.4 Semi-structured interviews with course participants Semi-structured interviews with BalancedView course participants were conducted to explore participant conceptualizations of weight stigma and to examine what they did and did not find influential or helpful within the course. Semi-structured interviews, with a topic-based interview guide were chosen, as such a flexible approach is useful in exploring phenomena about which little is known (Tolich & Davidson, 1999).  The interview guide was developed based on the literature review, preliminary findings from the development phase of BalancedView and collaborative discussions with the BalancedView project management team and my doctoral committee. The interview guide was tested in the pilot phase. Participants had favourable comments about the guide and no suggested changes. The interview guide is attached in Appendix C. The main topics on the guide were: contextual information about participant; reasons for participating; overall perceptions of the 91  course; changes as a result of the course; perceptions of course process and format; and perceptions of various stigma reduction strategies. Three interviews were conducted with participants in the pilot phase. The three interviewees in the pilot stage were all women. Ten more interviews occurred following implementation of BalancedView (two men and eight women).55 The results from the pilot interviews are not focused on in my findings here.  Knowing how many interviews was enough was a challenge in this study. As Thomson (2011) has stated, on average, theoretical saturation occurs with approximately 25 interviews, although may occur with as few as 10 or require as many as 30. In the case of BalancedView, I had the benefit of data triangulation and found that fewer interviews were necessary than initially thought to achieve theoretical saturation. I found that with the women in particular, I reached a point where I was not getting new information from the interviews and I surmised that it was time to stop. I wanted to continue to interview men, however, as I discuss later, there were so few men who had taken BalancedView (n = 8) that I was not able to recruit further men. Of the limited males who took the course, few responded to recruitment emails. One simply stated he was too busy to participate.  Due to geographic barriers associated with this being a provincial project, many participants were interviewed over the phone. Of the 10 interviews in the implementation phase, five interviews were over the phone, one occurred by email and four were in-person. Email, phone and in-person interviews each have benefits and drawbacks. Email interviews are convenient and provide an opportunity for the interviewer to craft well thought out questions and                                                 55 Details about the professions of interviewees are provided in the sample subsection of this chapter (section 3.7).  92  the interviewee to reflect on how to thoughtfully reply (Hamilton & Bowers, 2006). However, the interviews may also suffer from a loss of spontaneity and lack of verbal (or visual) cues (Hamilton & Bowers, 2006). Telephone interviews have the benefit of being cost efficient and a timely way to collect otherwise difficult to access data, as well as of allowing the interviewer to take notes unobtrusively (Novick, 2008). Above and beyond the email interview, telephone interviews also provide verbal social cues. However, as with the email interview, visual cues are not present (Opdenakker, 2006). A primary benefit of the in-person interview then is that visual social cues, such as body language, can provide the researcher with extra details. However, the tradeoff is that in-person interviewing suffers from challenges with access and cost. For instance, participants living in a different geographic area might be missed or, if accessed, this can be more costly and time consuming (Opdenakker, 2006). It is also important to note that whether an interview is more advantageous on the phone or in-person depends on the interviewee’s (and the interviewer’s) interaction preferences and styles (Gubrium & Holstein, 2001), which can unfortunately be difficult to know in advance of the interview.  3.5.5 Group interview and focus group with committee members  In addition to interviews with course participants, I also wanted to understand committee member perceptions concerning the ways in which weight stigma could be conceptualized. Although I had data on this from participant observation and document analysis, I wanted to corroborate and potentially expand on my interpretations of the former data set through additional qualitative work. With this in mind, I conducted one small group interview with two (female) committee members who were very involved from the outset (see interview guide in Appendix C). Through this, I solicited their perceptions of how weight stigma can be conceptualized and how to best address it. These two participants were selected as a result of 93  their regular and active involvement from the beginning of the study. I initially intended to conduct further interviews of this sort, but due to time constraints and the intention of the project team to use a focus group with committee members in 2015 as part of the evaluation, I decided to use the data from this focus group instead. I co-facilitated this focus group56 with the other contracted evaluator as part of the project evaluation. A benefit of a focus group (or group interview) is the opportunity for the researcher to witness participants react to one another, through which both diversity in perceptions and consensus can be highlighted. Focus groups are also a cost effective way of conducting research (Belzile & Öberg, 2012). A disadvantage of the focus group is that the emphasis is on group knowledge, rather than individual knowledge and not all individuals may feel comfortable expressing opinions that differ from the majority (Acocella, 2012).  The following questions, which were developed by myself and the other project evaluator, with input from the project managers, were used to guide the focus group:  • Process Evaluation Questions: o Did you feel engaged in the process (of developing BalancedView)? o What were the engagement strategies (e.g. meetings, email, document review, etc.) that work best for you? Why? o What would have contributed to more (better?) engagement? o We want to take a step back now and look at the whole project taking a collaborative approach. You have dedicated many hours to review materials, provide feedback and participate in meetings. How do you think the product is different (better?) because of that approach? o What makes the approach preferable (if it does) to a cheaper and faster strategy of                                                 56 The focus group lasted 70 minutes. 94  having a contractor do all the work? o Was any specific group or stakeholder missing from the conversations? What would they do differently? • Outcome Evaluation Questions: o How do you conceptualize weight stigma? o How has your understanding about weight stigma evolved as a result of participating in this process, if at all? o What did you learn about weight stigma as a result of participating in this process? 3.6 Recruitment For the pilot we aimed to recruit 25 health care providers to test the training and the pre-course and post-course questionnaires. Monetary incentives ($50 each) were offered. The team sent out one email that resulted in over 300 replies of interest. From this large group the project managers selected people to participate, based on an equal number of people from each health authority region, with an emphasis on diversity of type of health care job. Due to technical issues with the online platform during the pilot phase, only eight people were able to complete, despite two rounds of recruitment. To recruit among this pilot group for interviews, a member of the PHSA project team sent out an email invitation on my behalf. Though this method I completed three interviews.  After the pilot, once BalancedView was ready to be implemented, the project team provided incentives for a further 74 people to take the pre-course and post-course questionnaires, the course itself and the three- and six-month follow-up surveys. Participants for this cohort were selected in two ways. Firstly, we recruited from the formerly acquired list of interested pilot participants. Secondly, we recruited through the project managers who sent emails to their 95  networks asking individuals to fill out a fluid survey indicating their interest in participating in the longitudinal study and providing some basic demographic information. Out of all those interested, individuals were divided according to health authority region, with an aim of equal representation from each region. If there was not enough interest to achieve regional representation, the extra spots were equally distributed among the rest of the health authorities in BC. Previously interested individuals from the pilot stage were prioritized over those that filled out the fluid survey. In addition, these previously interested individuals were prioritized based on response time. Furthermore, individuals who filled out the fluid survey were prioritized based on occupation or regions that are typically under-represented among interested individuals (e.g. physicians, individuals in the North or individuals working with the First Nations Health Authority).  The other recruitment strategy beyond the longitudinal cohort was to develop a convenience sample (e.g. to target participants who were easy to reach). The course was publicly launched in Spring 2015 (after already being soft launched among the longitudinal evaluation cohort (n = 74) and potential participants were notified about it through online postings, emails sent through committee members and word of mouth. Anyone who identified as a health care provider was able to register and access the course.  To recruit for course interviewees I initially relied on the project management team to send out an invitation email to the longitudinal cohort on my behalf. A second recruitment email was later sent to the larger cohort on my behalf. This strategy yielded only women and no physicians, however, after which I targeted my recruitment towards these participants by email 96  with individually addressed invitations, to attempt to have a more varied sample. Next, the details about the course participants who met the inclusion criteria for this study are presented.  3.7 Sample This subsection provides details on the sample included in this study, including course participants and interviewees. As of September 2016, when data collection for this dissertation ended, BalancedView had been widely disseminated and implemented. A total of 925 registrants had signed up for BalancedView, including pilot participants. Of the larger group, some were excluded from the case study analysis. Exclusion criteria included: location outside of BC or Canada (thus out of scope); incomplete registration profiles leading to inability to assess for study eligibility; registrants with administrative roles in the BalancedView development; or registrants who identified as students (and thus not yet in practice). After excluding these participants, 654 participants remained, not including pilot participants. From this larger cohort, 249 participants completed the course in the implementation phase (38.01% completion rate). These 249 participants are the main focus of this study. From this group, 74 individuals participated in the longitudinal evaluation, described above, that was undertaken by the other evaluator and myself. A total of 56 participants from this sub-sample completed the three-month follow-up questionnaires and 46 completed the six-month follow-up questionnaires. Demographics for the participants who completed BalancedView (N = 249) are highlighted in the below tables. For additional details on the demographics of the longitudinal evaluation, please see Appendix D. As Appendix D shows, demographics were very similar for those who participated in follow-up.     97  Table 4 Gender, age and region of course participants (N = 249)  Demographics n  %    Gender (completers)57   Female/woman  240 96.39% Male/man 8 3.21% Other (i.e. 'both') 1 0.40%    Age (completers)58   18 - 24 8 3.2% 25 - 34 75 30.1% 35 - 44 61 24.5% 45 - 54 52 20.9% 55 - 64 31 12.4% 65 or above 6 2.4% Prefer not to answer 16 6.4%    Region in BC   Fraser region  24 9.64% Interior region  16 6.43% Vancouver Island region  19 7.63% Vancouver coastal region  174 69.88% Northern region  14 5.62% Unknown BC location  2 0.80%                                                    57 For inclusivity and social justice reasons, rather than have a drop-down list for gender with limited response options, we had an open field for text-based responses to allow for participants who may wish to identify outside the gender binary. Gender had a similar breakdown for all registrants, including those who did not complete. Among the 654 registrants, 93.27% identified as female or woman, 5.35% as male or man and 1.38% other (e.g. transman, queer, prefer not to answer).  58 Age distribution among all 654 registrants was similar: 3.5% age 18-24, 25.1% age 25-34, 26.8% age 35-44, 22.3% age 45-54, 15.9% age 55-64, 2.0% age 65+, 4.4% prefer not to answer. 98   Table 5 Profession of course participants (N = 249)  Profession n  %    Registered nurse 132 53.01% Dietitian 39 15.66% Counsellor 13 5.22% Administrative role  10 4.02% Other59  9 3.61% Dental profession (e.g. dentist or hygienist) 8 3.21% Mental health worker  7 2.81% Social worker 6 2.41% Registered psychiatric nurse 5 2.01% Occupational therapist 5 2.01% Speech and hearing health professional 5 2.01% Physical therapist 2 0.80% Physician 2 0.80% Public health 2 0.80% Recreational therapist 2 0.80% Audiometric technician 2 0.80% As the first table above shows, the majority of participants who took BalancedView identified as ‘female’ or as ‘women’. One substantial limitation is that very few men (n = 8) completed the BalancedView course. The results are thus not theoretically generalizable to men. Only one person completed the course who identified as being outside of the gender binary (they identified as ‘both’). The results therefore do not speak to the utility of this course among those who identify as a gender other than man or woman. While very few men signed up for the course the problem was not an issue of completion among men, as much as an issue of registration. For a number of somewhat unknown reasons it was women who signed up for BalancedView. One                                                 59 ‘Other’ includes professions that were only mentioned one time each, like hearing and vision screener, sonographer and psychologist. 99  possible reason for this is that women may have been drawn to this topic because weight stigma is often considered a gendered issue (Fikkan & Rothblum, 2012). When examining the professional breakdown of BalancedView participants, over half were nurses, which is a female dominated profession, perhaps explaining in part why the gender breakdown is skewed towards women. The second most common profession was dietetics. Only two physicians completed the course. The results thus also do not speak to the utility of this course among physicians.60 Next, I provide details about the interviewee sample. The interview sample included the following participants shown in the below table. The details provided here are also used to identify interviewee quotes throughout the thesis.61  Table 6 Interviewee details (n = 10) Gender Job details Interview medium   Female Social worker, eating disorders In-person  Female Dietitian, chronic disease Phone Female Administrative role, weight management Phone  Female Registered nurse, eating disorders Phone  Female Registered psychiatric nurse, community mental health Email  Female Community nurse In-person Female  Dietitian, chronic disease In-person Female Social worker, community mental health Phone Male Technician and supervisor, diagnostic imaging Phone Male Dietitian and food policy In-person                                                   60 It is unknown why few physicians signed up for the course. Possible reasons include, but are not limited to: busy schedules or preferences for learning modalities that are shorter and/or in person. This is followed up on in the discussion chapter (chapter six).  61 In addition to these ten interviewees, the pilot interview sample comprised of the following: female, healthy living coordinator and counsellor, phone interviewee; female, dental hygienist in public health, phone interviewee; female, dietitian weight management and eating disorders, phone interviewee. 100  3.8 Analysis  The analytic techniques for case studies that Yin (2003) overviews are primarily related to ‘pattern matching’, where researchers strive to compare their predicted theories or, ‘patterns’ with empirical data. Since I began this study with an exploratory lens, moving to the explanatory stage later, I began with ‘pattern identification’, or identifying empirical patterns, rather than matching predicted theories to patterns in the data, although as the study progressed the analysis increasingly became an iterative back and forth (abductive) approach between data-driven and theory-driven analysis. Pattern matching and pattern identification cohere with the qualitative analytic techniques of thematic analysis (Braun & Clarke, 2006; Clarke & Braun, 2013) and document analysis, which also involves organizing data into themes (Bowen, 2009). For my interview and focus group data I drew on the steps of thematic analysis outlined by Braun and Clarke (2006) and Clarke and Braun (2013), as follows: 1) read or review transcripts/other data sources and think about preliminary codes; 2) tag data with preliminary codes; 3) combine codes into preliminary themes and think about how to describe each theme; 4) theorize how themes explain data and revisit data if themes are incomplete; 5) formally define themes, which should be more than simply re-occurring patterns and should in some way answer the research questions; and 6) write report and ‘member check’ to see if themes have consensus among participants. This specific analytic strategy was chosen for analysing my transcripts as it lays out a clear process in which analysis can unfold, thus promoting the trustworthiness of the research. It also has the benefit of being theoretically flexible and relevant to multiple research paradigms (Clarke & Braun, 2013). 101  For my document analysis I grounded my work in Bowen’s (2009) conceptualization of the analytic process in document analysis:  The analytic procedure entails finding, selecting, appraising (making sense of), and synthesizing data contained in documents. Document analysis yields data – excerpts, quotations or entire passages – that are then organized into major themes. (Bowen, 2009, p 28) In conducting this document analysis, I again drew on Braun and Clarke’s (2006) stepwise approach. For, as noted by Bowen (2009), the latter part of document analysis involves thematic analysis. However, as Bowen (2009) suggested, unlike in thematic analysis, the first stage of document analysis is to conduct a “. . . first-pass document review” to ascertain which portions of the text are relevant to the research questions and should be analyzed thematically. Thus, for my document analysis, I combined steps one and two from Braun and Clarke (2006), in that reading and initial coding were done at once and then coded portions were re-read and analyzed further. This was necessary as many of the background documents from this study that were used in the document analysis included extra information that was above and beyond the research questions.  To code the transcripts and other documents I relied on a combination of coding manually in Microsoft Word (used for documents from the development phase and for transcript data) and coding in NVivo for qualitative data exported from the online platform of BalancedView for the implementation phase (i.e. scripted reflections and qualitative questionnaire comments). The reason for this two fold technique was to retain the benefits of being ‘close’ rather than ‘distant’ from the data that is associated with manual coding, while also 102  benefitting from the speed and simplicity of computer assisted software (Welsh, 2002), which was especially helpful for the lengthy 450 page qualitative export from the BalancedView course and questionnaires.  Quantitative data from the pre-course and post-course questionnaires were analyzed using SPSS Version 20.62 I checked for normality of variables using quantile-quantile (Q-Q) plots, histograms, skewness, kurtosis statistics and statistical tests for normality. I ran a Cronbach’s Alpha reliability analysis on any scales used at both the pre-course and post-course measurement points that had mean scores in order to assess for how closely related given items in a scale were. Descriptive statistics such as frequencies and mean are presented. I performed a paired sample t-test on attitudinal scales used in the pre-course questionnaire and post-course questionnaire. Quantitative data from the three- and six-month follow-up questionnaires were analyzed as part of the BalancedView evaluation. In this study I present descriptive statistics for the main outcome variables of interest from three- and six-month follow-up. More details on long-term changes are available in Appendix D, which contains the evaluation report undertaken by myself and the other evaluation consultant. As mentioned, one case study criterion I used was member checking. Thus, in the analysis phase, steering and advisory committee members participated in a meeting where I presented an overview interim results in Spring 2016. Committee members had an opportunity to ask questions or comment. No issues were raised concerning the preliminary interpretations.                                                 62 As part of data cleaning, data were screened for extreme scores outside of reference ranges and missing variables. Data were also screened to ensure only eligible participants were included. Data were originally exported in three separate Excel files and transferred to SPSS. Once this transfer was complete, data in SPSS were cross-referenced to the Excel files as a double-checking measure to ensure no errors had been made.  103  However, it is also important to note that the level of participation from BalancedView stakeholders was much greater in the development phase of the course and when deciding on what methods to use than in the analysis phase. As such, the findings presented should be understood primarily as my interpretation of the data.  3.8.1 Reflexivity  Bourdieu (1992) noted that social science is inherently prone to ‘bias’ and that researchers should thus strive to be reflective of their preconceptions and assumptions and how these influence interpretations of data. Reflexivity requires researchers to position and locate themselves within texts and is part of the reflective turn within social science (Lea, 2013). It is an iterative process where researchers seek to understand the influence of their own assumptions, experiences and social locations on the research process and outcomes (Harris & White, 2013). For Bourdieu, reflexivity was intended to increase objectivity (Knafo, 2016). However, it could be argued that social science will never be objective and, perhaps, nor should it need to strive to be. Instead, however, engaging in a reflexive analysis of one’s own influence on the work may be used to “. . . enhance the trustworthiness, transparency and accountability” of the research (Finlay, 2002, p 211). Through reflexivity, researchers and authors address how they came to write the text and how their subjectivity informed the work. In doing so, they demonstrate self-awareness so that readers can make judgements about the author’s perspective (Richardson, 2000). In my study, reflexive consideration of the influence I had on my research was thus important. I came to this research with a pre-existing opinion on weight stigma. Initially, I aligned myself strongly with the perspectives of fat acceptance advocates arguing that medical frames on 104  weight were harmful, hence I was committed to advancing this argument in relation to weight stigma. As I progressed through the study, I increasingly realized the value of being more open minded, as coupling my insights with those of others can only strengthen our understanding of how to address weight stigma and responses to it. I also came to reflect on how my own past experiences and embodiment had impacted how I related to weight stigma and this project.  As a teenager, like many young girls in current society, I experienced weight bias. As a dancer who physically matured early, I was occasionally bullied for my weight in my earlier teen years. This experience shaped how I related to issues of weight as a teen and young adult. There was a period of time when I, like many young women, struggled with body image. I spent a great deal of time feeling that my body was too large, trying to change it and blaming myself for not being able to achieve (for any length of time) the thin ideal I had internalized and set as a goal. As I moved into adulthood I was able to learn to let go of these ideals and to accept my natural body size and shape. My process of accepting my body was a multi-faceted journey, however, it was in part fuelled by learning that one’s weight was not as personally malleable as I had been led to believe and that ideas of fat as bad and unhealthy were based on social constructions. I learned to no longer blame myself (or others) for weight gain and let go of my internalized stigmatizing beliefs. In turn I felt an amazing sense of freedom and improved wellness. This personal journey has informed my perspective on weight stigma. I believe that because I agonized over my body so much as a teen and young adult and have experienced some of the things I write about that I have some insight into weight stigma. I aimed to use this insight to help me interpret the data gathered, while also unpacking the personal from other perspectives.  105  That said, I also recognize that I have thin privilege,63 which also affects how I relate to weight stigma. My body is relatively culturally normative and others most likely perceive it as both ‘healthy’ and ‘normal’. The result of this is that, despite my best intentions, I realize I may sometimes find myself able to look the other way in the face of weight stigma, rather than challenge it. During my study I attempted to keep in mind how my thin privilege affected my ideas and the actions. Of note, reflexivity was more deeply embedded into the first stage of my study during the development of BalancedView. This is because the development stage was an embodied experience where I was often physically present with participants and subjectivity played a large part in the analysis, particularly around the role of emotions. After the implementation of BalancedView, however, I was mostly working with text and numeric data. I felt more removed from the research process and participants and, given the emphasis on quantitative analysis during this stage, more required to adopt analytic techniques with a greater level of objectivity. Next ethics are discussed.  3.9 Ethics  This study has been reviewed and approved by the University of British Columbia (UBC) Behavioural Research Ethics Board (BREB) as minimal risk. In accordance with UBC BREB ethics policies, the following were approved as part of this study: recruitment scripts for the pilot test, course and interviews; interview schedules/guides; questionnaires; and informed consent forms for the pilot test, course and questionnaires and interviews. A waiver of consent was                                                 63 Thin privilege refers to “. . . the unearned advantages conferred to thinner people. It is a key pathway through which fat oppression is maintained” (Bacon, O’Reilly & Aphramor, p 42). It also is something that, much like white privilege, remains largely invisible to those with this privilege and it is “. . . common for privilege to remain below our radar” Bacon, O’Reilly & Aphramor, p 45). This, thin privilege may affect how people who are thin relate to the topics of weight, fatness and weight stigma.  106  acquired from UBC ethics for the focus group, as it was initially intended as part of the evaluation (and thus verbally consented to) and posed minimal risk to participants (no identifying details were used in the research). Below I discuss ethical issues that I grappled with in this study.   From an ethical perspective, the principle of ongoing, informed, signed consent is the gold standard.64 There were a few deviations in this study from standard consent processes, in addition to the waiver of focus group consent mentioned above. For the interviews, I solicited signed informed consent. Since the course itself was online, consent was an online process. If participants did not want to participate in the research they did not need to participate in the course. Ethically, this did not initially present any issues, since the course was not required to be taken and was intended to be voluntary. However, later on, in the analysis stage, I discovered that some organizations had mandated that their employees take the course – as per participant self-report – which raised ethical issues around voluntary participation in research and freedom to withdraw. This is mitigated by the instructions in the consent form where participants could withdraw from the study by contacting the research team. I did not receive any withdrawals. However, should this have been the case, the participants would still have been able to access the course content, but would have had their data excluded from the research study.  Another deviation from signed informed consent concerned participant observation. For my participant observation with the committees, rather than solicit signed consent, I solicited verbal consent on multiple occasions through the project and conceptualized consent as a                                                 64 As per the Tri-council policy statement: Ethical conduct for research involving humans (Natural Sciences and Engineering Research Council of Canada, Social Sciences and Humanities Research Council of Canada, Canadian Institutes of Health Research, Canadian Institutes of Health Research, & desLibris – Documents, 2010). 107  process. Wiles, Heath, Crow and Charles (2005) suggest that one potential alternative to signed consent at the initiation of an observation project is to conceptualize consent as a process where researchers balance participant needs and rights (e.g. adequate information provided to participants or protection from harm) with important research objectives (e.g. access to normally difficult to access data). Consent then becomes something that occurs over time rather than at the outset. Consent as a process can be helpful particularly in ethnographic types of observatory research where it is not always known at the outset exactly what consent is needed for or what the focus of the observation will be (Wiles et al., 2005), as was the case in my observation. As part of this process I verbally provided committee participants with my intention to collect data on process through my attendance in meetings and encouraged them to contact me or the project managers if they had questions or concerns. Committee participants were also provided with written information regarding the process of data collection. I additionally received a written letter of approval from the PHSA to conduct my process-related research and conduct observation, which was enclosed with my ethics application.  Within this ‘consent as process’ conceptualization, I remained alert to any ethical issues that would necessitate signed consent for observation. I took Moore and Savage’s (2002) recommendation on observation and consent and, rather than try to pre-determine what consent would look like, was instead responsive to ethical issues that arose as the project unfolded. In the end, avoiding signed consent for observation was possible because I solicited regular verbal consent, was committed to fully protecting the privacy and confidentiality of anything particular individuals said or did and the study was low risk, with no foreseeable harm to participants with the level of anonymity I was providing.  108  In the next two chapters I summarize my findings. The first chapter focuses on research question one and the second on research question two. In each findings chapter, the findings from the development phase are discussed first, followed by findings from the implementation of BalancedView. This is because, despite many commonalities, the findings for each question varied between these two phases of the case study.   109  Chapter 4: Findings regarding research question one / Conceptualizing weight stigma  This chapter focuses on findings from my first research question (‘what are the different ways that weight stigma in health care can be conceptualized?’). I first reflect on how weight stigma was conceptualized in the development stage of BalancedView. I then discuss how participants who took BalancedView perceived weight stigma.  4.1 How was weight stigma conceptualized during the development of BalancedView? As discussed in the methodology chapter, I conducted a thematic and document analysis of the data collected in the development phase of BalancedView to explore how stakeholders involved in developing the course conceptualized weight stigma. In due course, I present these thematic findings. However, first I encourage readers to delve deeper into the story of the development of BalancedView. This story is a prerequisite to understanding the themes that follow, as the development of BalancedView spanned two years and over that time there were great shifts in how weight stigma was conceptualized.  4.1.1 A story of shifting conceptualizations in BalancedView’s development – an interlude  Before and after  Related, I did.  Her pain became my crusade.  Contested I was. (Research journal, 2015) The development phase of BalancedView was a complicated and lengthy process during which emotions ran high and the essence of weight stigma was debated, often hotly. As the above poem from my research journal suggests, my perspective on weight stigma was challenged during the development of BalancedView by other stakeholders involved. I began my 110  involvement in BalancedView with an interest in advancing the social justice agenda on behalf of people who experienced weight stigma and challenging the medicalization of weight. However, this (radical) perspective, largely learned through fat acceptance circles, was not well accepted by some individuals involved in developing the course. From the inception of BalancedView through to the start of the pilot test there were notable shifts in how I and others involved in the project related to the notion of weight stigma.  The story of this project began long before funding was secured. In 2010, the Province of BC was actively working to develop an ‘obesity reduction strategy’. Simultaneously, and in the years following, several health care leaders in BC began to advocate for a shift away from weight-centred approaches to health and instead focus on healthy behaviours irrespective of body weight (Provincial Health Services Authority, 2011). The rationale for this was that a focus on weight and weight loss as a measure of health were shown to cause harms such as disordered eating, weight cycling and stigma. In 2011, I was invited to sit on a newly formed committee sponsored by BC Mental Health and Substance Use Services. This committee – the Promoting Healthy Weights Working Group (PHWWG) – consisted of a group of interdisciplinary health professionals with a mission to prevent disordered eating and reduce weight stigma. In its early days, there was an emphasis on shifting the focus from weight to ‘health for all,’ regardless of size. These efforts were aligned with a Health At Every Size agenda. The work of this committee and health care leaders in BC and elsewhere in Canada with similar perspectives, among other factors, ultimately led to the PHSA commissioning a review of the evidence on weight, health and weight stigma. In their final report from this review, the PHSA (2013b) argued that it was time for a ‘paradigm shift’ away from weight-focused health promotion and towards health promotion emphasizing wellbeing. One rationale for this was that weight-centred health care 111  may contribute to rather than reduce weight stigma. Notably, this approach was a radical shift from the previous obesity reduction strategy led also by the PHSA. The political climate around the topics of weight and health, while charged, was shifting. Shortly after this report was released, in fiscal year 2012/2013, BC Mental Health and Substance Use Services secured funding for BalancedView. This emergent way of thinking about weight, health and weight stigma informed BalancedView’s initial development. For my part, I entered this project with a very strong opinion of what caused weight stigma and what should be done to address it. Like some others involved, I was convinced that medicalizing weight was a central contributing factor to weight stigma and that dismantling the medicalization of obesity was needed. My emotions about the issue ran high and I saw the current weight centric focus in health care as a social justice issue. I (mostly) believed the dominant research on weight and health to be ‘factually incorrect’ and contested the notion that weight-centred approaches to health were helpful and, in fact, saw these approaches as both harmful and discriminatory (O’Reilly & Sixsmith, 2012). I believed that a shift towards Health At Every Size (HAES) and the contestation of the common belief that fat equates to unhealthy and weight loss to healthy would be a step in resolving weight stigma. One of the early activities of the BalancedView team was to undertake a scoping review to explore the problem of weight stigma and how to address it. This involved a systematic literature review and interviews conducted by two consultants/contractors, with input from the committees. Daníelsdóttir et al. (2010) had already reported that there was surprisingly little research on the subject of weight bias reduction or evidence on what might actually work to address this stigma. The systematic literature review in 2013 led by the two contractors reiterated 112  this point: “The bottom line . . . is that we really don’t know what’s going on, what works to decrease prejudice or why, in part because we still have a relatively poor understanding of prejudice” (PHSA, 2013a, p 28). Despite this lack of clarity, ‘labelling’ (Link and Phelan, 2001) higher weights as a health issue was identified as a component of the stigma process.  The scoping review also involved 22 interviews with an international group of researchers and professionals who worked in areas related to weight stigma (n = 19) as well as patients who identified themselves as heavy and had experienced weight stigma within the BC health care system (n = 3). The findings were summarized in a lengthy report (see PHSA, 2013a) that was collaboratively developed by the two contractors, with input from members of the committees (including myself and the project managers).  The report strove to understand why weight-related bias exists and how it could be addressed (PHSA, 2013a). It thus provided me with valuable insight into how weight stigma was conceptualized in the early days of developing the course, particularly since the process of developing the report involved regular meetings with the committees to get feedback. The contractors noted that potential causal theories of weight stigma included attribution theory and social consensus. They suggested that interventions to reduce weight stigma should draw from these theories and more, particularly since the evidence suggested that multi-prong approaches to weight stigma reduction were most effective (PHSA, 2013a). Notably, the final scoping review also laid out the presumed relationship between medicalization and weight stigma:  Another contributing but relatively underexplored factor may be the medicalization of overweight and obesity in Western society – that is, the “labeling of all fat people with the medical labels of ‘overweight’ and ‘obesity’ and presuming that all people falling 113  within these categories are inherently unhealthy” (personal communication, C. O’Reilly, February, 2013). In this line of thinking, if people with overweight or obesity need to be cured, then something is wrong with them. Or, as Beausoleil and Ward (2009) observe, “The association of fat with poor health has translated into a fear of fat within the population and subsequent disdain for those who are different, who do not fit the desired norm” (p 1). This process may indeed be one in which a powerful group (medicine/health) in society plays a role in determining who is “different” and thus subject to discrimination. (PHSA, 2013a, p 18)  Given the potential link between medicalizing weight and weight stigma, the contractors proposed that weight bias reduction in health care should include a section on “Myth Busters - Providing evidence about weight, weight bias and health” (PHSA, 2013a, p 5). They also argued for the importance of other components to the course, including providing an opportunity for self-reflection, exposure to the experience of being heavy and using opinion leaders to influence thinking.  These recommendations were agreed upon by members of the committees and further developed by the content contractor. This content contractor, as discussed in the methodology case description, worked with the committees, the project managers and I to develop a lengthy Word document with the proposed curriculum content. Over several months this document – which contained multiple modules, including one on medicalization – was finalized and again consensus among committee members was seemingly reached. The interview quote below illustrates the process of working together to achieve consensus within the development phase: 114  But just, yeah, making sure that people felt heard and that they felt like their feedback was being incorporated into the resource, right. Like, if people feel like they’re giving . . . feedback and voicing things but there not being any change as a result, then that’s frustrating. And so we really have to make sure that-- yeah, and I think that was probably the most work . . . like, all that process of taking in all of that feedback . . . and making sure that we incorporated it in a way that people were going to feel happy with it. And I think we got there, but it definitely took a lot of effort, a lot of time, more meetings than we anticipated . . . (Committee interviewee, female)  The initial intent for the 3-year project was to develop the course within the first half of the funding stage and to spend the rest of the time pilot testing and implementing the course across BC. Ultimately, the development phase was much more drawn out than anticipated, due in part to the contentious nature of the topic matter.   Between December 2013 and the spring of 2014, it became clear to all on the committees that the initial consensus that we must challenge the medicalization of weight to reduce weight stigma in health care was not actually a consensus. Some people involved in the project rejected these ideas and had emotional reactions to them. Consider the following quote from a committee interviewee:  One of the challenges was really getting to a place where everyone could agree with the approach that we were taking. I think that took a lot of conversation, a lot of dialogue, as you know. And we’ve talked about earlier-- so that was a significant challenge, I think, and just in terms of process, I think that the reason that conversation happened so late is because we started out with a content that was a lot… And I feel like people didn’t really 115  read it at first. So when it became something like a script that people actually could go through and digest, that’s when the feedback started coming out. So I think that delayed us a little bit just because we had to do so many revisions late in the game. But they were necessary, so I think the outcome was really good. So that was probably one of the major challenges. (Committee interviewee, female)   In December 2013, a daylong meeting was held to discuss the proposed draft content of the course, which had been previously agreed upon and translated into a mock-up of what it would look like online. The intent of the meeting was to review the visual presentation of the material from a technical perspective. However, hours of the meeting were instead devoted to discussing content, specifically pertaining to the medicalization module. This was the first time that divergent opinions really emerged about the extent to which fatness should be seen as a health issue and whether assuming it is a health issue is stigmatizing or not. Specifically, some committee members perceived some of the language around medicalization to be too strong, off-putting and inflammatory, with one public health professional going on to argue that the proposed framing was ‘offensive’ and that obesity was a serious health issue.  The project leads attempted to redirect the conversation, arguing that this meeting was about the technical and visual flow of information for the online resource and not about content, and that we could play with ‘language’ later. In the next few months, however, despite some ‘toning down’ of language, the conflicts over the medicalization issue continued and it became increasingly clear that a few committee members, as well as some external stakeholders, were so opposed to an anti-medicalization agenda that they would potentially attempt to stop the course from proceeding if it continued as intended. It seemed that some of the health care stakeholders 116  involved in this project were not ready, after all, for a ‘paradigm shift’. As one committee member shared, they thought that all the positive work being pursued by the BalancedView team would be undone if the project tried to deny the relationship between overweight and obesity and mortality and morbidity. This committee member felt that it was incorrect to contend, as module three initially did, that there was only questionable or weak evidence to support that “[m]ortality rates increase with increasing degrees of overweight, as measured by BMI”.65 In their perspective, the literature is very clear that overweight and obesity are both significant risk factors for early death and increased morbidity. Further, this participant felt that attempting to deny this risk would render pointless any efforts to address the obesogenic environment.  Behind the scenes, the situation became more difficult, with project managers and sponsors receiving more and more complaints about the seemingly too radical and ‘factually incorrect’ approach that was being taken with respect to challenging beliefs about fatness as unhealthy. As one particularly oppositional individual decreed  – who also happened to hold a leadership position within health care in BC – “I will die on this hill”, implying that they would go to great lengths to defend the perspective that obesity is bad for your health and prevent the course from articulating anything to the contrary.  Ultimately, a political decision was made. In order to move forward with implementation we had to revise the content more significantly. Through a series of smaller meetings and several more committee meetings, we revised the material. Much like the name of the resource, we opted to present a ‘balanced view’ on the medicalization issue. In the module where the ‘myths’ about                                                 65 BC Mental Health and Substance Use Services (2013, December). 117  body weight were initially challenged, we instead presented the two sides of the debate on medicalization, with the aim of giving learners enough information to make the decision themselves. Although I was angry and upset by the initial conversations about the need for this shift, by the time of the pilot I was able to accept this strategy and could see some potential benefits. In reflecting on my position on weight, health and weight stigma at the time of the development of BalancedView I was able to see how my feelings connected to my own past struggles with my body image and weight. This is shown in the below text from my research journal: Given my struggles with my own weight and body image, when I learned about HAES, fat acceptance and weight stigma, I instinctively gravitated towards what some might characterize as a more “radical” perspective on these topic. I aligned myself with fat acceptance activists and fat studies scholars with a HAES orientation. I eagerly accepted the messages of writers like Bacon and Aphramor (2011) who argued that weight focused approaches to health are based on flawed science and cause harm. After all, I had experienced the inaccuracies and harms from such assumptions myself. I contested the notion that fat equates to unhealthy and weight loss is healthy or sustainable. I became an avid reader of fat acceptance blogs, where it was common to hear fat activists write about their own experiences with dieting, weight loss and disordered eating. An oft-heard refrain on these blogs was that focusing on weight is harmful and moralizing. I valued these personal stories and readily identified with them. As someone with a keen social justice interest, the moralizing nature of weight-centred health became a crusade to me. All of this culminated in my master’s thesis, in which I was . . . passionately committed 118  to radically changing the public health climate on weight and health in BC. In BalancedView, this was manifested by strongly advocating during the committee meetings for an anti-medicalization agenda. (Research journal)  Looking back, I see that my position on weight, health and weight stigma was affected by my own emotional experiences. The fact that I had personally experienced the emotional repercussions from living in a weight-centred culture helped me understand the harms of said culture. As well, the process of recognizing how my own emotions and experiences influenced my thinking led me to develop a curiosity about how others’ emotions and experiences might influence how they thought about weight and weight bias. Specifically, I wondered how might the perspectives of those who supported medicalizing obesity be influenced by their own emotions? It was notable that the people in this project who were strongly opposed to challenging the ‘myths’ about weight and health were those whose careers were invested in maintaining those strongly held cultural beliefs. Thus, perhaps a sense of defensiveness contributed to some of the reactions of committee members. Those opposed to the staunch anti-medicalization perspective initially taken seemed to be, more often than not, male and/or educated as physicians. Privilege, both male privilege and educational privilege were thus also dynamics that I was curious about.  With this background in mind, I now delve into the themes from the development of BalancedView pertaining to how weight stigma was conceptualized. Themes were as follows:  1) Stigma as process 2) Stigma as attitude/belief/stereotype 3) Stigma as discriminatory behaviour or outcome 4) Stigma as causally complex 119  5) Stigma as related to medicalization… or not?: A divisive issue 6) Stigma as related to emotions Each is discussed in turn. I begin with stigma as process.  4.1.2 Stigma as process  The first theme from the development phase pertains to the essence of weight stigma in health care as conceptualized by those of us involved in the development of BalancedView. The various individuals involved in developing BalancedView characterized weight stigma as a problem beginning with biased attitudes and/or beliefs that lead to discriminatory behaviours (or actions or inactions) and resultant consequences for those who are the subjects of weight stigma; in other words, as a process, with multiple interrelated components. Extensive discussion was held at the committee meetings to clarify that while weight stigma can be a problem for people across the weight spectrum, it disproportionately affects individuals classified as overweight or obese or who visually appear heavy. A focus group participant (male) also pointed out how weight stigma intersects with other social issues like racism and misogyny:  So my conceptualization of weight stigma is just another in-- you know, homophobia or racism, misogyny. It’s just like-- and the interactions and complexity with all of those things, right . . . But it’s just another one of those things, another one of those reasons to treat people badly, and there’s intersections with a whole bunch of other reasons to treat people badly.  Much like in the broader literature, the terms weight bias, prejudice and stigma were often used interchangeably by those involved in this project. However, despite the terminology used, there was agreement that the problem was a process encapsulating one or more of the following: negative attitudes and beliefs about weight, that often lead to discriminatory treatment 120  of heavier people and adverse outcomes for those who experience weight stigma. All participants in the development phase saw the problem as beginning with attitudes or judgments about heavier people and this was linked to varying extents with the other components. Those actively involved in the course development tended to conceptualize the problem as a closely coupled issue of attitudes/stereotypes that result in negative outcomes for the stigmatized. For example, consider the following quote from an interview with a committee member:  Well, I would define weight stigma as the devaluing of a person based on their weight, whether they’re overweight or underweight. And in terms of the issues that it presents, I think another, you know, link that we have into this piece is the mental health piece. And so the mental health concerns that can come along with being weight stigmatized are really important to us. And also just, you know, inequitable care. I think that people who are of a certain size can experience inequitable healthcare, and that can lead to health problems like difficulties with health seeking . . . (Committee interviewee, female) The course itself reflected a similar perspective and positioned weight stigma as a process involving an intersection of several components: attitudes/beliefs/stereotypes and discriminatory behaviour or outcome. Each of these related components is discussed as follows as a theme in its own right. 4.1.3 Stigma as negative attitude/belief/stereotypes  In the scoping review phase of the course development, PHSA (2013a) described the attitudinal component of this process as follows:  Weight bias - negative weight-related attitudes, beliefs, assumptions and judgments toward individuals who are overweight and obese . . . These attitudes are often 121  manifested by false and negative stereotypes which cast overweight and/or obese individuals as being physically unattractive, incompetent, lazy, unmotivated, less competent, non-compliant, lacking self-discipline, and sloppy (Puhl & Heuer, 2009; Rukavina & Li, 2008). (p 14)  An interviewee involved in the committees spoke similarly of this aspect of the process, connecting it again to inequitable outcomes:  Weight bias [involves] the assumptions that people are making based on a person’s appearance, right. And so-- and the fact that when that person happens to be a bigger person, then those assumptions tend to be negative, right. And so from the literature we know that there’s a lot of negative characteristics that healthcare professionals and the general population-- but healthcare professionals as well are not excluded from that in making those assumptions about people. And that, that in itself is problematic because it does affect the way that they approach that person, things that they’ve already kind of made up in their mind about who this person is and what their story is, without even having asked those questions. And so I think that-- yeah, I think that that can contribute to . . . the inequitable care that they might receive, in terms of the fact that they might experience, you know, more mental health concerns. Or might be less likely to-- or other types of concerns as well because they might be less likely to access care and that that’s extremely problematic. (Committee interviewee, female)   While it was readily apparent that negative attitudes about overweight and obesity were seen as a key part of weight stigma by participants in this project, it was less immediately obvious what ‘counted’ as a weight-biased attitude. Early on in the development phase of BalancedView, the 122  project team noted that one of their key objectives from the project was to create “Improved attitudes towards weight” (PHSA, 2013, August, p 3). The conversation then turned to defining what was meant by attitudes about weight. In the process of developing the project’s outcomes measurement framework (OMF), the other evaluator and I were tasked with reviewing the literature on weight stigmatizing attitudes and their measurement and bringing this information back to the committees. In doing so we conceptualized weight-biased attitudes in the OMF as follows: “. . . assumptions and reactions, implicit and explicit . . . Including attitudes and beliefs about: i) fatness as invariably unhealthy,66 ii) fatness as within personal control, iii) non-compliance with health care recommendations, iv) attitudes about treating obese patients” (PHSA, 2015, April, p 4). The above definition reflects the current literature on attitudes as involving an implicit (automatic and emotive) component, as well as an explicit (well thought out, aligned with beliefs) component (Watts & Cranney, 2009). 4.1.4 Stigma as discriminatory behaviour or outcome  It was generally agreed that the negative attitudes and assumptions held by health care providers about heavier people often led to discrimination and negative outcomes for overweight and obese patients. In the scoping review, PHSA (2013a) conceptualized the discrimination or behavioural component as:  Weight discrimination – “unequal, or unfair treatment of people because of their weight” (Puhl, n.d., pg.1). Thus, discrimination extends beyond beliefs and attitudes to unjust or unfair actions and behaviours toward people who are overweight or obese (Ciao & Latner                                                 66 While the team agreed to an OMF that articulated that one of the project aims was to reduce medicalized attitudes and beliefs about fatness as invariably unhealthy, the extent to which medicalization was seen as stigmatizing or necessary was an area where there was extensive debate, as discussed more shortly.   123  2011). Discrimination can take many forms, from verbal comments and derogatory remarks to excluding, avoiding, ignoring or rejecting, to cyber-bullying, physical aggression and victimization (Puhl, 2011). (PHSA, 2013a, p 14)  In the context of health care, discrimination was understood to include a refusal to assess or treat patients who are heavier, along with declarations by health care providers that health problems experienced by overweight or obese people must be attributable to their weight. One of the patients who was video recorded on film for the BalancedView course described this experience:   I would define weight bias as, you know, when you meet a person, whether they are big or small, presuming that their body size can tell you about their lifestyle, about their history, about their priorities, about how they spend their days and nights, when in fact you really can’t . . . For me the stigma could be defined as the negative ways you get treated based on those presumptions that people have.  Discrimination was also seen to occur when health care equipment was unable to accommodate the needs of fat patients. One patient interviewed on film as part of BalancedView discussed this problem poignantly:  Several months ago I broke my ankle and I ended up in the hospital, I needed to have surgery on it. And in advance of that I had been to the doctor, to my family doctor a few times and I had asked to – I was interested in knowing what my blood pressure was because I knew that being a bit heavier meant that I might have some risk associated with high blood pressure and they weren’t able to find a high blood pressure cuff that could get a reading of my blood pressure. Then I went into the hospital and, um, they tried to take my blood pressure, but no one could really seem to figure out how to take it or 124  where there could be a cuff that they could use and so I ended up in surgery and I had like this extreme like crisis high blood pressure which, um, professionally upset the anesthesiologist and the surgeon, um, and they felt that I shouldn’t have even been brought to surgery with blood pressure that high. No one had . . . taken any ownership of taking care of the fact that I needed to have my blood pressure taken just as a patient who’s going into surgery. I think what I got out of that health care encounter if you will is that it wasn’t just a matter of one person saying, “Oh it doesn’t matter to me whether you have your blood pressure taken” it was that the hospital, whoever orders the blood pressure cuffs doesn’t have a sense that they need to get any for someone who’s got a really big arm. Um, that at all these different levels of decision making they’re not taking into account this kind of patient who they presume and, in fact, probably really believe needs and deserves the best treatment that they can provide. But the way that that comes out in practice… is that you are not receiving the same care as someone else as someone who is smaller.  Participants involved in course development also described weight stigma as a process culminating in negative health outcomes for heavier people, including poor mental health, barriers to accessing health care and disordered eating. As one interviewee involved in the committees shared:  And so the mental health concerns that can come along with being weight stigmatized are really important to us. And also just, you know, inequitable care. I think that people who are of a certain size can experience inequitable healthcare, and that can lead to health problems like difficulties with health seeking. (Committee interviewee, female)  125  As was articulated through the scoping review:  . . . poor body image, low self-esteem, low self-confidence, loneliness, sense of self-worthlessness, depression, anxiety and other psychological disorders, suicidal thoughts and acts, maladaptive eating patterns and eating disorders, avoidance of physical activity, and stress-induced pathophysiology. (PHSA, 2013a, p 2)   The emphasis on these consequences was also reflected in the course itself, which provided evidence on how weight stigma may lead to unhealthy eating, doctors spending less time with patients or heavier patients delaying or forgoing medical care. One of the patients interviewed on film for BalancedView also described avoiding health care as a result of felt stigma:  Once when the swine flu was going around my- my friend had it and I had been feeling quite sick so I umm went to the doctor, a doctor that I didn’t know and I walked in and the doctor … kind of scoffed . . . [and] said, “Well I hope you’ve been tested for diabetes”. And, you know I had just had all my levels checked let’s say a couple months previous and when I told him that he persist[ed], wanting to talk about my body size in a really disrespectful way and as you know at clinics and ERs time is really of the essence and he kind of pushed my actual needs out of the schedule in order to talk about my body size. What that means for someone, or how that impacts you, is that like I think we think the issue is “Oh fat people you can’t take the tough love from any doctor” and actually it’s “Well I don’t want to go back because I’m not getting my needs met, like you did not address the reason I was there”.  In addition to conceptualizing weight stigma as a process of judgment and, ultimately, consequences, individuals involved in the development phase also talked about weight stigma as 126  a complex social issue with multiple social factors contributing to its existence. The contributing factors presumed to lead to weight stigma varied from person to person and were less well understood than the attitudinal or discrimination components. This is discussed in the below theme.  4.1.5 Stigma as causally complex  Those involved in the course development viewed weight-biased attitudes as resulting in part from the belief that fatness is controllable (attribution theory) and from the influence of pervasive cultural and social norms (social norms and consensus theory). While these were seen as part and parcel of weight bias, it was also emphasized that there was a lot about the complex problem of weight stigma that we simply do not understand:  Yeah. I think it’s difficult to get to the root. For sure. Because, you know, if you think about all the characteristics that are attached to someone who has a larger body size, like, where do those come from? The fact that-- if you think people who are bigger are unhealthy and that’s under your personal individual control, then-- yeah, where does that belief come from in terms of health? . . . So what contributes to that and what-- and, you know, how-- I don’t know. I feel like there can be so many reasons that that has happened. And part of them are cultural, part of them may be biological, I don’t know. (Committee interviewee, female)  Despite many unknowns about the origins of weight stigma – for example, the potential role of biology in contributing to bias – attempting to understand what causes weight stigma was seen as important during the development phase of the course. As stated in the PHSA (2013a) scoping review: “Any approach to change must address the fundamental causes of stigma” (p 12). With 127  this in mind, consideration of causal explanations for weight stigma occurred during the early stages of developing the course content. Explanations that were considered particularly plausible and that had ready agreement were based on attribution theory and social norms/consensus theory. Attribution theory, as discussed in the literature review, is a common causal explanation for weight stigma (Puhl & Brownell, 2003). Data from the development phase supported the notion that perceptions of fatness as something controllable that fat people should be blamed for contribute to weight bias. As laid out by the content contractor in the content document:  Consistent with Link and Phelan’s (2001) principle of getting at the root causes of stigma, Crandall and colleagues have argued that attributions about fatness come from a connected set of convictions, beliefs and values that form a belief system or ideology. Many different values and beliefs are correlated with weight bias, including right-wing authoritarianism, the Protestant work Ethic (strong emphasis on self control, and the view that hard work and determination yield success), and conservative political ideology, for example. All of these have in common the notion that people are responsible for what happens and that they deserve what they get (Crandall & Reser, 2005). This emphasis on personal responsibility is the foundation of Western individualism, which gets to the core of weight bias. (BC Mental Health and Substance Use Services, 2013, May, p 6).  Committee discussions at the table echoed this sentiment, without delving as deeply into the connection between individualistic values and the meritocracy. As I captured in a journal entry “people at the table seemed to think it was due in part to misinformation . . . not being aware that weight is not as controllable as we think”. From the above we can see that attribution theory 128  provides two insights into why weight stigma exists: one being the (misguided) perception that weight is entirely controllable; and the other being a sense of righteousness deriving from the Western values of individualism and merit.   Social norms around weight were also seen as part of contributing to weight stigma amongst health care providers. We live in a society where thinness is constructed as the most attractive and socially desirable and fatness as less desirable aesthetically and socially. Health care providers, as discussed by committee members, are not immune to the influence of these social norms. As one committee interviewee discussed: “I think part of it has to do with our ideals of beauty and what is attractive and what’s not attractive”. Weight bias is also something that health care providers, like most members of society, are likely to internalize at a young age due to how older members of the population talk about fatness:  But today, I just think it’s perpetuated because we do, like, adults do have those ideas. And very quickly kids pick that up, and internalize those same ideals. But, again, I don’t-- I’m not really sure how it’s so internalized. (Committee interviewee, female) The content document also stipulated the same thoughts:  In Western society, the cultural norm for female beauty and attractiveness includes extreme slenderness, and from an early age, children are aware of the negative connotations of being overweight. It is likely, therefore, that by the time individuals reach adulthood, they have developed well-rehearsed and complex associative networks in memory between the concepts, “fatness” and “thinness” and negative and positive affective modes respectively. (BC Mental Health and Substance Use Services, 2013, May, p 6)  129  The above evidence illustrates the importance of social beauty norms about weight. But what about social consensus? Social consensus theory, as discussed in the literature review, builds on theories of social norms, adding that socially influential people are particularly likely to be able to influence the attitudes of others (Puhl, Schwartz & Brownell, 2005).  Social consensus was mostly discussed in the context of BalancedView’s development as a strategy for bias reduction. However, I also observed social consensus theory to play out in the group dynamics at the committee meetings insofar as the presence of influential health care leaders who were more mainstream obesity thinkers in respected professional roles shifted the tone of the conversation towards one that was arguably more biased. For instance, once one prominent obesity professional expressed their perspective that fatness was unhealthy, this seemed to give license for others to express similar beliefs, in a way that was metaphorically like the opening of floodgates. After this, more people spoke up with this same perspective and it appeared there was a risk of the funding not continuing if the course took a strong anti-medicalization stance. Social consensus also operated to silence the voices of those with opposing perspectives. On more than one occasion, a committee member came up to me after a meeting and thanked me for standing up for the ‘unbiased perspective’. They said that they did not feel safe speaking up once others expressed their pro-medicalization agenda.67 Thus, we see how some professionals involved in BalancedView had an active role in enabling some to express their views, while silencing others, thereby shaping group norms.                                                  67 To ensure people had an opportunity to express themselves if they did not feel comfortable speaking up at meetings, the project leads invited committee members to provide more feedback by email, if desired.  130  Interestingly, the power of social consensus within the group seemed to impact the expression of explicit attitudes but have less impact on changing implicit attitudes. When observing group dynamics in the committee meetings, when influential leaders at the table verbally advocated for HAES or an anti-medicalization agenda, this created a sense of consensus and ‘silenced’ those who perhaps were not entirely in agreement with this perspective. However, as became clear when other influential leaders opposed the medicalization perspective, the extent to which others’ attitudes had truly changed seemed limited, as the presence of a pro-medicalization, ‘obesity as unhealthy’ voice in essence gave permission to those who held similar beliefs to be more vocal about them. One committee interviewee described how social consensus influenced what was said, but not necessarily thought, during committee meetings. She described how “. . . sometimes there are people who do feel particularly strongly or who are a bit more dominant in conversations  . . . it tends to be taken over by, you know, those few voices that are the strongest”. She concluded by saying that after meetings, however, people would privately express different perspectives and speculated that people “. . .maybe didn’t feel comfortable” expressing their opinions openly (Committee interviewee, female). 4.1.6 Medicalization as stigmatizing… or not?: A divisive issue This theme pertains to the extent to which medicalization was seen as contributing to weight stigma during the development phase. As I show, there were diverse perspectives. Is medicalization stigmatizing or is it not? Those involved in the development of this project simply could not agree. There were three different ways that the medicalization debate was framed. Firstly, at the start of the project, there was a strong push to label medicalization as inaccurate and stigmatizing. For instance, in text prepared for the course in October of 2013, module three 131  planned to contest the following claims about obesity, based on the work of Campos, Saguy, Ernsberger, Oliver & Gaesser (2006). To quote the initial iteration of module three:  Claim #1: 'Almost all countries (high-income and low-income alike) are experiencing an obesity epidemic’ . . .  Claim #2: ‘Mortality rates increase with increasing degrees of overweight, as measured by BMI’ . . .  Claim #3: ‘The data linking overweight and obesity to adverse health outcomes are well established and undeniable’ . . .  Claim #4: ‘Significant long-term weight loss is a practical goal (is achievable), and will improve health’ . . . (BC Mental Health and Substance Use Services, 2013, October, p 33).68 Each claim was individually contested with evidence disputing it and references to other studies. Within this push to label medicalization as part of stigma the role of social consensus was discussed as contributing to why these claims persist despite evidence to the contrary. Consider an excerpt from the draft scripts for the course:   . . . If there truly is very little compelling evidence that losing weight is a positive health choice, why do weight-centred approaches prevail? Recent research on this very question turned up some fascinating answers to this question. In a study where the language that weight experts and policy makers typically use and rely upon was thoroughly examined, it was found that claims by health experts were taken as evidence even though those                                                 68 This is an excerpt from draft scripts for the course that drew on Campos et al. (2006). 132  experts did not substantiate the claims. (BC Mental Health and Substance Use Services, 2013, October, p 38) As this quote shows, people with ‘expert’ status may have power to influence the beliefs of other about weight, even if their beliefs are not backed up with evidence.  The second perspective taken by some in the group was to see the medicalization of weight as helpful, necessary and accurate, while also wanting to see weight bias decreased. For example, as one committee member argued, weight bias is a pervasive problem that needs to be corrected. However, this participant felt that efforts to address overweight and obesity are also required as overweight and obesity were seen to contribute to increasing morbidity and premature mortality.  The third point of view, and the one which many, if not most, committee members could compromise on, was that medicalizing weight has its pros and cons and that while weight may at times be related to health, an overemphasis on weight can also lead to more stigma, weight cycling, disordered eating and so on. This was the approach ultimately taken in the course, in which both the benefits and consequences of medicalization were presented. An interviewee from the committees also shared:  I feel like . . . it was important . . . that we present a balanced perspective. And I don’t know that I’m necessarily really radically on one side or the other. Like, I am kind of in that space of understanding both sides of the argument, but wanting to learn more. (Committee interviewee, female)   When examining why committee members either agreed with or rejected medicalization as contributing to weight stigma, personal and professional experiences appeared to come into 133  play. I observed that sometimes those in favour of medicalizing obesity were male or doctors, steeped in a particular way of thinking and epistemic tradition, while those with a more open perspective about the topic tended to be women with experience professionally in eating disorders. One committee member described the professional dynamics at play in how weight stigma was conceptualized, as shown in the following interview quote:   I think, you know, the culture that we’re working within is a medicalized culture but, you know . . . health professionals-- a lot of them are coming from that model and so . . . it’s just really difficult to question something that is so engrained in how you’ve been trained and how you just sort of practiced. And I think also physicians being kind of at the top of the hierarchy and being, you know, within a medicalized culture, that also comes to play, I think, in terms of their sort of dominance, in a way. (Committee interviewee, female) I believe those of us who felt an anti-medicalization perspective was the correct one were also influenced by their/our personal and professional experiences. Several worked in eating disorder treatment and thus – while I cannot claim cause and effect – had what I feel to be an enlightened perspective on the topic. At least two of us had past body image struggles. Regardless of why we believed what we did, in the pragmatic tradition, we agreed that we needed to stop fighting over what was true and instead focus on what we could do, in practical terms, with what we know to reduce weight stigma. The debate then becomes less about ‘truth’ and more about action. Next, the relation of emotions to weight stigma is discussed.  134  4.1.7 Stigma as related to emotions: Emotions matter to how weight stigma is conceptualized As Lea (2013) discusses, in qualitative research, exploration often leads to more questions than answers. This has very much been case here concerning the examination of emotions and weight stigma. I was surprised to observe that emotions were so central to the ways in which weight stigma was thought about and talked about by different people involved in BalancedView’s development. The development data show that emotions matter to how weight stigma was conceptualized. However, further investigation is required to understand this relationship well. A limitation of the development data and my subsequent analysis is that since I did not set out to collect data on this, I do not have enough data focused specifically around the topic of emotions. In my analysis, I noticed myself frequently tagging data with the code “emotions!”. While I had read about the relationship between emotions and stigma in the work of Link and colleagues (2004), I was initially not particularly convinced of its relevance and was more interested in the centrality of labelling and medicalization. Now, however, I perceive that further exploration of the subject of emotions and its relationship to weight bias is needed due to the clear emergence of this within the data. While emotions played a central part of how weight stigma was understood by those of us involved in the development of BalancedView, this was not obvious early on in the development phase and instead emerged in the latter stages of this portion of data collection. Below, I discuss the various ways I observed emotions exerting their influence in the early stages of my case study.   Firstly, in this study I noticed that the language used to discuss weight was seen to evoke an emotional reaction among health care professionals. Drawing on insights from discourse analysis (Jacobs, 2006), in framing social problems, the language that is used has implications 135  for our feelings and understandings of the issue. This is especially the case when language involves metaphors, as metaphors help to construct and shape how we think and feel about situations being discussed (O’Halloran, 2006). For example, in the content document originally created by the contractor it was discussed how terms like ‘combat’, ‘epidemic’, or ‘crisis’ tend to evoke a sense of panic about obesity. How obesity is talked about matters, some committee members articulated. When the common cultural discourse is one of an epidemic that costs taxpayers, it makes sense that experiences of fear at the former and unfairness at the latter could arise.  Secondly, among health care providers, where a common recommendation is for heavier patients to lose weight, another emotion that can arise is a sense of frustration if the client does not lose weight over time. Again, in the initial content document it was articulated that health care providers may experience feelings of frustration as a result of the perhaps mistaken perception that their patients are not following their directions. This sense of frustration connects directly back to attribution theory and the perception that body weight is personally controllable with appropriate effort. Thirdly, weight is something that in today’s society, many people have strong, deeply entrenched beliefs about, as the following interview quote suggests. Emotional reactions arose in the development of BalancedView when such beliefs were challenged. Yeah, I think that-- well, I think that part of the reason that that has been so contentious is because of some of those really strong engrained beliefs as well, right. I think that that makes it harder to be really open to accepting some of this kind of new, not new, but like to them maybe-- or to most people it’s new literature-- . . . that people honestly can’t really believe is true because they’ve been told the opposite for so long for, you know, by 136  so many different-- from so many different perspectives, both personally and in their, you know, professional careers. And I think that that’s just a really difficult step for people to take, to make that leap to just being really open to that. (Committee interviewee, female)  In the case of BalancedView, the deeply entrenched nature of the belief about fat as unhealthy was seen through sentiments such as, “I will die on this hill” (committee member). Some believe that getting fat people to lose at least some weight is imperative and are unwilling to compromise on this. Health care providers may be particularly invested in some of these beliefs about weight and health, due to the type of science they are exposed to in their education, coupled with a desire to promote health.   In my observations, when beliefs about weight, health and weight loss were discussed at committee meetings, emotions often ran high among health care providers, including those for and against medicalizing obesity. Others also noticed this emotional response, as this interview quote shows:  Just ‘cause we were getting-- we were having these discussions in meetings where we, again, were having these emotional reactions between people that were on one side or the other of feeling this way. (Committee interviewee, female)  Some other committee members also told me or told the group about the emotional reactions they were having during these challenging discussions. For example, one individual provided written feedback that hearing that weight loss interventions were stigmatizing felt shaming and dismissive of the genuine care of individuals like her involved in obesity management. In this case, when her professional choices were challenged and labelled as 137  stigmatizing, she experienced an emotional reaction. This emotional reaction then shaped how she conceptualized weight stigma.  It was common to witness heightened and palpable emotional states among committee members when debating over the supposedly true relationship between weight and health. This was seen through things like quickened speech, tone and use of terms like ‘offensive’ in response to perspectives different from one’s own. As an example, at one point in time my early position – that medicalization was harmful and discriminatory – was labelled as offensive by some committee members with divergent opinions. In hearing rebuttals to my opinion, I also felt upset at the perceived injustice, as I wrote in my research journal in 2013: Recently I was told that my master’s thesis in which I criticized the prevailing weight science and policy of the day as being inflammatory and based on biased, as opposed to legitimate “science” or “expertise” was “offensive”. I was told that my calling out weight science and policy as inflammatory or questioning the expertise of professionals in this field was in and of itself inflammatory . . . Maybe they were right. In fact, they are definitely right. I DID offend them. They, like all others, are entitled to feel and label their emotions. And yes, it WAS inflammatory, insofar as it challenged deeply held and internalized norms of the time. But does this make it bad? Does this mean I should stop speaking on the subject? No. Because if science, coupled with the prevailing norms of the time, is being used to . . . enforce an inequitable social order that has very real consequences for people, I intend to call that [out].  In concluding this subsection, given the elusive nature of writing about emotions, particularly those of other people, coupled with the fact that I did not intend to collect data on 138  emotions, many questions remain about emotions and weight stigma. What emotions relate to weight bias and why? Disgust, for example, has previously been hypothesized to be part of weight bias (O’Brien, Daníelsdóttir, Ólafsson, Hansdóttir, Fridjónsdóttir & Jónsdóttir, 2013). Other issues at play may include: righteousness, healthism, negative feelings about one’s own body and defensiveness and a need to protect one’s professional identity. Regarding this last point, it is notable that individuals on the committees who at times expressed beliefs about fat being unhealthy or controllable, also tended to be people who were professionally invested in some way in such viewpoints (e.g. had careers where obesity management was a focus). One potential reason for this (of many) could be a sense of defensiveness about one’s professional choices and a need for self-protection of sorts. These are issues for future consideration. Next, how weight stigma was conceptualized by those who took the course in the implementation phase is discussed.  4.2 How was weight stigma conceptualized by participants during the implementation of BalancedView?  Next, I build on understandings of weight stigma and its conceptualizations from the development phase and explore how health care providers who took the course viewed weight stigma and, when relevant, highlight areas of agreement and disagreement. Based on a thematic analysis of the qualitative data collected from participants who took the course and their perceptions of weight stigma, I discuss the following themes:  1) Weight stigma as a process involving attitudes/beliefs and actions/unfair treatment 2) Controllability beliefs as stigmatizing 3) Medicalization: More harmful than helpful?  139  Each is discussed below.  4.2.1 Weight stigma as a process involving attitudes/beliefs and actions/unfair treatment In the implementation phase, the previous conceptualization of stigma as a process involving attitudes/beliefs/stereotypes leading to unfair treatment based on weight was once again present. I present evidence of this only briefly here as it is coheres with the already well-articulated theme of the same described in the prior subsection. As one interview participant articulated, weight stigma involves both treatment (or an action component) and thoughts or judgements:  . . . treating somebody differently because of their weight that you maybe wouldn’t treat other people. So whether that [means] judging them based on their weight or thinking that they can or can’t do things because of their weight. Making serious judgments about their health because of their weight. All of that would be [inaudible] weight stigma. (Female, administrative role, weight management, phone interviewee) Ultimately, data from the implementation phase underscored the previously discussed notion that weight stigma is about prejudgments we make about people based on implicit and explicit attitudes and beliefs that lead us to treat people in a prejudicial or discriminatory way, whether intentional or not. For example, as another interviewee discussed, it involves both the judgements and assumptions and the subsequent way you interact with people, based on weight:  I personally define weight stigma as when you judge a person based on how they look and by look I mean being fat or skinny, so to speak, the layman’s terms. And then it’s like it’s how you interact with them. It’s what you think about them. It’s things that you 140  say. And-- yeah, making assumptions about who they are based on their weight. (Female, social worker, eating disorders, in-person interviewee) Next, I return to the question of what attitudes and beliefs were conceptualized as part of weight stigma. When looking at what attitudes and beliefs were viewed as connected to weight stigma in the implementation phase, the main patterns were around beliefs about fatness as controllable as contributing to stigma and also beliefs about fatness as a disease as being potentially stigmatizing. A small number of participants also discussed biased attitudes as extending beyond these two dimensions, including, for example, beliefs around attractiveness that are deeply culturally engrained. However, these were emphasized less often and were less salient to the health care context. The controllability and disease dimensions of weight stigmatizing attitudes are discussed next as themes, based on perceptions of course participants.  4.2.2 Controllability beliefs as stigmatizing  As is consistent with attribution theory, health care providers who took BalancedView articulated a relationship between a belief in the controllability of weight and weight bias. As one participant shared in a course reflection: “By thinking about weight as a product of choices, it allows some people to feel very self-righteous” (ID 442, course reflection). One interviewee, for instance when discussing how she conceptualized weight stigma referred to judgements about the controllability of weight and the subsequent moral implications people derive from these assumptions:  It’s when people make judgments right away based on someone’s weight, and usually those are negative judgments. Obese people are less [inaudible]. Obese people are lazy, they’re gluttonous. They eat too much. They can’t control their weight. Those ideas of 141  blaming the victim . . . It [inaudible] sort of equate[s] fat with the moralistic idea of being a bad person. (Female, dietitian, chronic disease, phone interviewee) While the data from the implementation phase suggested health care providers were able to acknowledge that controllability beliefs were a part of weight stigma in health care, it is also worth noting that this was a less evident pattern in the qualitative data from the BalancedView course than the potential stigmatizing nature of medicalizing obesity, insofar as it was brought up markedly less often.69 However, this discrepancy is likely explained by the fact that participants were directly asked after watching the medicalization debate and viewing the medicalization Prezi to provide their comments in a scripted reflection, thus prompting them to reflect on and share their thoughts on medicalization more deeply.  4.2.3 Disease beliefs: Medicalization as more harmful than helpful? Much as in the development phase of data collection, not all participants who took BalancedView perceived the medicalization of obesity to be part of weight stigma. However, a majority of participants, after exposure to the course content on the pros and cons of medicalization, either: 1) articulated significant stigma arising from medicalization (approximately 42%); or 2) were able to see some potential harms of medicalization, yet also some possible benefits (approximately 41%) (e.g. they could appreciate both sides of the ‘debate’). Far fewer participants were entirely in favour of medicalizing obesity or saw only                                                 69 There were approximately 10 pages of coded text regarding controllability beliefs, in comparison to 34 pages of coded text regarding medicalization.  142  positives in it (approximately 17%).70 Evidence of each of these different perspectives is shown below.  4.2.3.1 Medicalization as harmful and stigmatizing One common position taken in the scripted reflections after the medicalization module was to view the medicalization of obesity as problematic, leading to some combination of the following: shame for those whose bodies are pathologized; eating disorders and mental health issues; an overly simplistic view of weight and health; reinforcing or creating stigma; and as distracting for health care providers from other health problems that may be affecting the patient, among other potential issues. There were so many interesting and noteworthy quotes on this that it was difficult to choose what to highlight. I provide some brief glimpses into the richness of this data below.  Regarding the potentially shaming nature of labelling obesity as a disease, one participant shared in a scripted reflection:  My patients have expressed to me that they have experienced a lot of shame, depression, feelings of wanting to give up and, after receiving a diagnosis of overweight or obesity, [that] they are often not supported with the tools to address the diagnosis, but are told a weight range they must strive for. (ID 239, course reflection)  Others talked about the negative impacts of medicalization on patients in other terms, commenting on this pathologization negatively impacting emotional state and sense of self, for example:                                                   70 These percentages represent the proportion of the time each of these perspectives was coded in the implementation data.  143  . . . I have a friend who was recently told she was morbidly obese. She exercises regularly, all her blood work and other health indicators are within “normal” limits. This label has deeply saddened her - she already feels the stigma from her peer and professional group and now she feels it from her own family doctor [with] whom she has a great connection . . . This has had a[n] overall negative impact on her sense of self and her positive outlook on life has changed. (ID 753, course reflection)  The perceived negative impacts of medicalization also extended to disordered eating and related mental health issues. For example, as one participant shared in a course reflection: “I fear that there are some serious ramifications if we medicalize obesity. In my opinion this leads to disordered eating and mental health concerns” (ID 986, course reflection). Others talked about how medicalizing weight leads to overly simplistic thinking along with consequences like patients feeling ‘othered’ and ashamed, which in turn can lead to poor health habits and distress: I would fall on the side of the debate that sees medicalization as contributing to weight bias and stigmatization. Any time something is deemed “normal” or “outside of the normal range” we are falling into dichotomous thinking, or black and white thinking. The consequence of this is that people who don’t fall within the “normal” range are made to feel other-ed, which I think brings shame and a downward spiral into being less healthy, both physically and emotionally. We need to question where the “normal” category came from. Who deemed it so? What research was done to back it up? (ID 225, course reflection) 144  The consequences of assuming that obesity is defacto a disease, simply based on BMI, are too simplistic, some argued, creating ideas about the need to ‘fix’ the so-called problem, when there may not be a problem in the first place: I think that obesity, defined as a BMI of 30 or greater, is such a simplistic idea (simply a factor related to height and weight), that I cannot seriously see it as a disease. What is the definition of “disease?” It must be the opposite of “100% well” and must include some challenges to living a full life. I think there are so many examples of people with high BMI who are well and leading full lives that the idea of them having a disease is hard to take seriously. So, even if a patient is referred because of “obesity” I may not have viewed them as having a problem, aside from the obvious problem that is being created with the label. But because they are labelled as having a problem, then it must have a fix, and knowing that weight loss is often not realistic, how to productively move forward? (ID 203, course reflection) Much like the position taken in the early iteration of the BalancedView curriculum, some participants also discussed the process of labelling fatness as a medical concern and how those labels themselves were creating or reinforcing stigma. As one participant stated: Placing a label [on] something will only reinforce the stigma and I think that we still have a long way to go to shift the focus from treating obesity as a disease and having ideal goals or “normal” for people to achieve without first taking into account all of the other factors that are going on. (ID 688, course reflection)   145  Other participants saw these labels as problematic specifically because of the assumptions embedded within them, for example, that there is a ‘normal’ range that everyone should strive for and that all heavier people are unhealthy.  I have to agree that if you medicalize the terms “overweight ” and “obese” you are left with the assumption that there is a “normal” and that everyone should attempt to achieve this. I agree that using the BMI alone is ineffective in determining “good” health . . . Someone in the high range of the BMI may be in perfect health while a person in the “normal” BMI range may have a host of health problems. (ID 220, course reflection) A few participants went so far as to argue that medicalizing obesity was misguided in a similar way to the medicalization of breastfeeding or was akin to the historical tendency to consider certain races or homosexuality as diseased. Regarding this latter point, one participant stated:  [D]eclaring obesity a disease is contradictory. We say that many medical issues are higher in people of certain races, yet we do not classify those races as a disease (we once did) - and yet for some people weight may contribute negatively to their health and so we classify it as a disease? It is as she [the expert oppose