Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

How brand hatred shapes consumer perceptions and preferences Johannes, Boegershausen 2019

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

Item Metadata

Download

Media
24-ubc_2019_may_boegershausen_johannes.pdf [ 1.71MB ]
Metadata
JSON: 24-1.0378066.json
JSON-LD: 24-1.0378066-ld.json
RDF/XML (Pretty): 24-1.0378066-rdf.xml
RDF/JSON: 24-1.0378066-rdf.json
Turtle: 24-1.0378066-turtle.txt
N-Triples: 24-1.0378066-rdf-ntriples.txt
Original Record: 24-1.0378066-source.json
Full Text
24-1.0378066-fulltext.txt
Citation
24-1.0378066.ris

Full Text

       HOW BRAND HATRED SHAPES  CONSUMER PERCEPTIONS AND PREFERENCES  by  JOHANNES BOEGERSHAUSEN B.Sc., Catholic University of Eichstätt-Ingolstadt, 2009 M.Sc. (cum laude), Maastricht University, 2011    A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR	  OF	  PHILOSOPHY	  	  	  in	    THE	  FACULTY	  OF	  GRADUATE	  AND	  POSTDOCTORAL	  STUDIES	  	  (Business	  Administration)	  	  	  	  THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)   April 2019   © Johannes Boegershausen, 2019  ii The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled:  How Brand Hatred Shapes Consumer Perceptions and Preferences  submitted by Johannes Boegershausen in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Business Administration  Examining Committee: JoAndrea Hoegg, Marketing and Behavioural Science Co-supervisor Darren W. Dahl, Marketing and Behavioural Science Co-supervisor  Katherine White, Marketing and Behavioural Science Supervisory Committee Member Andrew Barron, Psychology University Examiner John C Ries, Strategy and Business Economics University Examiner  Additional Supervisory Committee Members: Steven Heine, Psychology Supervisory Committee Member      iii Abstract Consumers frequently experience negative feelings toward brands; yet existing research has predominantly focused on positive engagements with brands. The current dissertation examines one of the most extreme negative feelings—hatred—and explores how hatred for one brand affects competing brands. Many managers seem to believe that consumers’ hatred for a close competitor would not be harmful or might even be beneficial for their brand. Nine studies using qualitative, experimental, and field data demonstrate that in contrast to managers’ beliefs, hatred for a brand leads consumers to eschew close competitors from the same subcategory. Importantly, such preference shifts do not emerge when consumers are indifferent or dissatisfied. As feeling mistreated and exploited is central to hatred, it triggers concerns about self-protection, which results in avoiding close competitors. Several moderators (i.e., greater variance in consumer ratings and the usage of safety-inducing marketing cues) supporting the self-protection based account are identified. Taken together, this dissertation emphasizes that consumer relationships with brands do not operate in a vacuum. Documenting predictable shifts in preferences for competitors as a consequence of hatred for a brand underscores the importance of extending frameworks of consumer-brand connections to incorporate negative connections and account for effects beyond a focal brand. The findings of this dissertation also suggest that hatred – at least in a consumption context – tends to prompt individuals to be primarily concerned about protecting the self rather than annihilating the hated object (i.e., brand). When consumers experience hatred, self-protection concerns appear to be pivotal in shaping their preferences for subsequent consumption unless alternative motives might interfere and produce different preferences.     iv Lay Summary Many consumers harbor negative feelings toward firms and brands. This dissertation examines one of the most extreme negative feelings—hatred—and explores how hatred for one brand affects other competing brands. Past research suggests that consumers are likely to avoid hated brands. However, it less clear how hatred for one brand might affect preferences for other competing brands. Does hatred benefit or harm close competitors of a hated firm? According to managerial wisdom, consumers’ hatred for a close competitor would not be harmful or might even be beneficial for their brand. This dissertation demonstrates that such beliefs are largely misguided, as hatred for a brand prompts consumers to eschew close competitors from the same subcategory. As feeling mistreated and exploited is central to developing hatred for a brand, it makes consumers concerned about protecting the self, which in turn results in avoiding close competitors of the hated brand.           v Preface I am the primary author of the work presented in this Ph.D. dissertation. I was responsible for conducting the literature review, developing the hypotheses, designing the studies, collecting the data, analyzing the data, and preparing the manuscript. Additional contributions are described below. Data for the manager survey (see chapter 1) was collected by Chuck Howard, JoAndrea Hoegg, and Darren W. Dahl. The data from Yelp used in study 1 (chapter 5) was collected by Yijun Chen under my direction.  JoAndrea Hoegg, Darren W. Dahl, and Anne-Kathrin Klesse provided intellectual contributions and assisted in designing the experiments.  Ethical approval for all studies was obtained from UBC Office of Research Services Behavioural Review Board (Human Ethics) under the certificate H14-02966. Details on sample sizes and sampling procedures are provided in chapter 3 of this dissertation.        vi Table of Contents  Abstract .......................................................................................................................................... iii	  Lay Summary ................................................................................................................................. iv	  Preface ............................................................................................................................................. v	  Table of Contents ........................................................................................................................... vi	  List of Tables .................................................................................................................................. x	  List of Figures ................................................................................................................................ xi	  Acknowledgements ....................................................................................................................... xii	  Dedication ..................................................................................................................................... xv	  Chapter 1:	   Introduction ............................................................................................................. 1	  Chapter 2:	   Theoretical Development ........................................................................................ 5	  2.1	   Hatred Activates Self-Protection Concerns ....................................................................... 9	  2.2	   When Hatred Leads to Preference for a Same-Subcategory Brand: An Opportunity for Revenge ..................................................................................................................................... 12	  Chapter 3:	   Overview of Studies .............................................................................................. 13	  Chapter 4:	   Qualitative Study .................................................................................................. 14	  4.1	   Data Analysis ................................................................................................................... 15	  4.2	   Discussion ........................................................................................................................ 21	  Chapter 5:	   Study 1 .................................................................................................................. 22	  5.1	   Data .................................................................................................................................. 23	  5.2	   Results .............................................................................................................................. 25	  5.3	   Discussion ........................................................................................................................ 28	    vii Chapter 6:	   Study 2 .................................................................................................................. 29	  6.1	   Method ............................................................................................................................. 29	  6.2	   Results .............................................................................................................................. 30	  6.3	   Discussion ........................................................................................................................ 30	  Chapter 7:	   Study 3 .................................................................................................................. 32	  7.1	   Method ............................................................................................................................. 32	  7.2	   Results .............................................................................................................................. 34	  7.3	   Discussion ........................................................................................................................ 35	  Chapter 8:	   Study 4a ................................................................................................................ 36	  8.1	   Method ............................................................................................................................. 36	  8.2	   Results .............................................................................................................................. 38	  8.3	   Discussion ........................................................................................................................ 42	  Chapter 9:	   Study 4b ................................................................................................................ 43	  9.1	   Method ............................................................................................................................. 43	  9.2	   Results .............................................................................................................................. 44	  9.3	   Discussion ........................................................................................................................ 47	  Chapter 10:	   Study 5 .................................................................................................................. 48	  10.1	   Method ........................................................................................................................... 49	  10.2	   Results ............................................................................................................................ 51	  10.3	   Discussion ...................................................................................................................... 53	  Chapter 11:	   Study 6 .................................................................................................................. 54	  11.1	   Method ........................................................................................................................... 54	  11.2	   Results ............................................................................................................................ 55	    viii 11.3	   Discussion ...................................................................................................................... 57	  Chapter 12:	   Study 7 .................................................................................................................. 59	  12.1	   Method ........................................................................................................................... 60	  12.2	   Results ............................................................................................................................ 62	  12.3	   Discussion ...................................................................................................................... 63	  Chapter 13:	   General Discussion ............................................................................................... 64	  13.1	   Theoretical Contributions .............................................................................................. 64	  13.2	   Managerial Implications ................................................................................................ 68	  13.3	   Future Research and Concluding Remarks .................................................................... 70	  References ..................................................................................................................................... 76	  Appendices .................................................................................................................................... 92	  A1. Brand Hatred Prevalence Study ........................................................................................ 92	  A2. Interview Outline (Qualitative Study) ............................................................................... 93	  A3. Overview of Informants (Qualitative Study) .................................................................... 97	  A4. Cuisine as a Category Marker  (Study 1) ........................................................................ 100	  A5. Review- and Reviewer-Specific Control Variables (Study 1) ........................................ 101	  A6. Robustness Checks (Study 1) .......................................................................................... 102	  A7. Multiverse Analysis (Study 1) ......................................................................................... 104	  A8. Stimuli and Consumption Scenarios (Study 2) ............................................................... 105	  A9. Tax Advisor Scenario and Dependent Variable (Study 3) .............................................. 107	  A10. Tax Advisor Pretest (Study 3) ....................................................................................... 111	  A11. Hatred Manipulation Check (Studies 3 – 7) .................................................................. 112	  A12. Hotel Categorization Pretest (Study 4a) ........................................................................ 113	    ix A13. Hotel Scenario (Studies 4-6) ......................................................................................... 114	  A14. Overview of the Consumers’ Choice Set (Studies 4a, 4b, & 6) .................................... 116	  A15. Brand Recall Question (Study 4a) ................................................................................. 117	  A16. Competitor Strategies (Study 6) .................................................................................... 118	  A17. Consumers’ Self-Protection Concerns (Study 6) .......................................................... 119	  A18. Telecommunication Scenario (Study 7) ........................................................................ 121	  A19. Hatred Versus Anger Follow-Up Study ........................................................................ 124	      x List of Tables Table 1. Yelp Data Descriptive Statistics ..................................................................................... 25	  Table 2. Propensity and Timing of Consumers’ Subcategory Return .......................................... 26	      xi List of Figures Figure 1. Purchase Likelihood for Chain and Independent Hotels (Study 4a) ............................. 40	  Figure 2. Moderated Mediation Model (Study 4a) ....................................................................... 41	  Figure 3. Variance Manipulation (Study 5) .................................................................................. 50	  Figure 4. Purchase Likelihood for Chain and Independent Hotels (Study 5) ............................... 52	  Figure 5. The Effect of Different Intervention Strategies on Consumer Preferences (Study 6) ... 56	       xii Acknowledgements  Reaching this academic milestone would not have been possible without the advice, support, and encouragement of the incredible people around me. First and foremost, I am deeply grateful to my two advisors, Darren Dahl and JoAndrea Hoegg. Thank you both for all that you have done for me over the years. You went significantly beyond the call of duty to facilitate my professional and personal growth.   Darren, you were my first contact at UBC and you truly put Sauder’s motto “opening doors” into practice in our relationship. I do not think that I can even recount all the ways in which you helped me in my career and development. Thank you for taking a chance on me when few would have and for all your help in the final stages of the process. Your intermittent prep talks, tough love, and excellent advice on research and life matters were instrumental to the completion of my dissertation and for furthering my identity as a researcher.   Joey, words cannot adequately describe all the ways in which you made my time in Vancouver better. You are a true role model in all aspects. I learned countless small and big things about how to conduct research from you, but even more importantly, you instilled the persistence in me to ensure that my ideas come to fruition. Your steady presence, mentoring, and encouragement have made an enormous difference not just for me, but my entire family. Few individuals can rival your blend of being caring, brilliant, dedicated, and humorous. I will always be grateful for all that you have done for me and I aspire to one day be an advisor like the one you were to me.     xiii I would also like to explicitly thank my two wonderful dissertation committee members, Kate White and Steve Heine. Kate, thank you for always being extremely supportive and for the countless ideas and inspirations for designing clever experiments during my time at UBC. Your feedback and mentoring has been instrumental in sharpening my thinking and writing. Steve, I have always greatly enjoyed our conversations over the years and your input during the process of completing this dissertation was extremely valuable.  One of the things that I am going to miss immensely after graduating is the wonderful group of people that form the department of Marketing and Behavioural Science at UBC. Every single (faculty) member has made a positive contribution to my development during my time here. Thank you all for the wonderful conversations, experiences, and friendship during these formative years of my career. Yann Cornil has always provided a cheeky sense of humor and many inspiring conversations. I am also indebted to Karl Aquino not just for teaching me ways to ask interesting research questions and for numerous thought-provoking exchanges, but also for being a family (together with Nezihe) away from home for us.  My graduate school experience at UBC would not have been the same without the presence of my fellow Ph.D. students. Sina, thank you for inviting me to live with you, your quirky perspective on the world has brightened up many days. The friendship that developed out of being roommates – with your entire family – seems well positioned to stand the test of time. Thank you to all the other students from the old guard (especially Kirk, Thomas, and Sky), the more recent generations, and all the visiting students for your friendship and support. Chuck, thank you for being a good sport, connoisseur, and friend over the years. Rishad, I cherish the many friendly reviews as well as our thought-provoking conversations. Julian and Navid, you   xiv made the final stretch a lot more fun, thank you for the numerous moments of entertainment and enlightenment.  There are many other individuals who made a difference over the last years. I was fortunate to share an office with Anne Klesse. The collaboration and friendship that blossomed out this has made a difference for me. Thank you, Anne, for all your support and efforts over the years. Hannes Datta, thank you for all your thoughtful advice and being there to listen when I needed it most. Having met by happenstance, Abhishek Borah turned out to be an amazing friend and brilliant co-author. I also have many fond memories of sharing an office with Alexander Henkel and Robert Ciuchita. Thank for the numerous priceless moments and your friendship. Finally, merci to Benoit Aubert and Bernd Stauss for opening the world of academia to me.  Very special thanks to my parents, family and friends, who supported me especially when the going got tough. To my parents, I do not express this often enough, but I would not be where I am today without you. You gave me the means to realize my dreams and have always believed in me – Danke. Many thanks also to my godfather, Lutz Hoffmann, and his wife Mechthild, for their unwavering support throughout my life.   Finally, the most important person in my life over the last decade has been you, Suma. Thank you for not simply putting up with, but rather fully embracing my academic pursuits. Moving to the other side of the globe into an ocean of uncertainty is not an easy feat, yet, honestly, it seems to be one of the smaller things that you have done for me over the years. Having you in my life has made me a better person and I cannot imagine life without you. Life with you and our son Samuel is a gift that keeps on giving.     xv Dedication In loving memory of my grandparents:  Elisabeth and Paul Funk, and Maria and Leo Bögershausen.   1 Chapter 1: Introduction “Der Hasser lehrt uns immer wehrhaft bleiben.”  [“The hater teaches us to be on guard at all times.”]  Johann Wolfgang von Goethe (1803; The Natural Daughter)   Consumers frequently experience negative feelings toward brands. Indeed, negative feelings might even be more common than positive ones. In a recent study, 55% of respondents reported primarily negative feelings toward brands, including worry, dissatisfaction, and hatred (Fournier and Alvarez 2013; Romani, Grappi, and Dalli 2012). Although hatred—perhaps the most negative of all feelings toward brands—is often highlighted in the news (e.g., Stebbins et al. 2018) and in discussion forums and blogs (e.g., “I hate Starbucks” or “I hate Walmart” on Reddit), the extent to which consumers actually feel hatred toward brands is unclear. Is brand hatred widespread or is it just the rare individual that experiences such negative feelings toward brands? To gain insight into the prevalence of brand hatred in the marketplace, I asked 1,053 participants on Amazon Mechanical Turk (61.9% female, Mage = 35.7) whether they hated any brands. Participants responded to the following question: Are there any brands that you hate? (“Yes” vs. “No”). If they responded with “Yes”, they were prompted to list up to five hated brands. More than half of those surveyed (54%) reported hating at least one brand. The types of brands listed by consumers varied considerably, with the most frequently listed firms being from the apparel (18%), service (18%), consumer goods (12%), and technology (10%) sectors (see appendix A1 for additional details).  Brand hatred appears to be relatively common, but are companies aware that consumers harbor such negative feelings? An advertising campaign by Spirit Airlines suggests that they are.   2 In 2014, Spirit invited consumers to visit its website and express their hatred for Spirit or other airline brands, suggesting that they understand that customers do indeed experience hatred. Interestingly, at the end of the campaign Spirit marketers rejoiced that "60% of the total hate [of the almost 30,000 consumers who participated] went to other airlines" (Spirit 2014). Presumably, Spirit’s managers believed that consumers’ hatred for competitors would be beneficial for their brand. Such a belief appears to be consistent with managerial wisdom. Specifically, in a brief survey, 60 managers sampled from a variety of industries (e.g., restaurants, retail, automotive, consulting, travel and tourism) indicated whether they thought that their brand would profit, suffer, or be unaffected if consumers hated one of their close competitors. The majority indicated that hatred toward a close competitor would not cause their brand to suffer. Indeed, 38% believed it would be beneficial and 37% believed it would not affect their brand. This begs the question: do companies really remain unaffected or even benefit from hatred directed at their close competitors? The objective of the current dissertation is to answer this question. Specifically, I examine how hatred toward one brand influences consumer preferences toward competing brands. Brand hatred is characterized by "intense feelings of dislike, animosity, hostility and aversion" (Matsumoto 2009, p. 230), and typically stems from the perception that the hated other (e.g., person, brand) imposes material and psychological harm (Gervais and Fessler 2017; Sternberg 2003). When consumers feel hatred for a brand, they experience concerns about the potential for further harm and negative outcomes from other related or similar sources (e.g., other members of a hated social group; Sternberg and Sternberg 2008). My central prediction in the context of brand choice is that after a hatred-evoking consumption experience with a focal brand (e.g., a boutique hotel named Hotel Abri) consumers are not only unlikely to go back to   3 the focal brand (i.e., Hotel Abri), but will also eschew other close competitors (i.e., other boutique hotels) that belong to the same-subcategory as the focal brand. In order to protect themselves from similar aversive experiences (Griskevicius and Kenrick 2013), consumers will instead opt for a brand from a different subcategory such as a chain hotel (e.g., Hilton).  Building on the assertion in Goethe’s opening quote that the existence of people experiencing hatred should make others vigilant, I propose that hatred for a brand activates self-protection concerns in those experiencing hatred as well. These concerns then in turn lead consumers to prefer competitors from a different subcategory to close competitors from the same subcategory. To test these predictions I employ a multi-method approach using qualitative, correlational, field, and experimental data. By demonstrating the effect of brand hatred on consumer preferences for competing brands I make several contributions of theoretical as well as practical significance. First and foremost, I identify the potential of brand hatred to exert effects beyond avoidance of the focal object of these feelings. Specifically, brand hatred is shown to reliably shift consumer preferences for competing brands in choice sets that do not even feature the hated brand. Second, I contribute to work on consumer-brand connections by identifying self-protection concerns as a driver of consumer preferences when feelings of hatred are activated. Potentially, my account of how hatred triggers self-protection concerns can deepen understanding of how actions and corresponding perceptions of a focal brand affect other brands as demonstrated in research on innovations, scandals, crises, and trust violations (e.g., Borah and Tellis 2016; Darke, Ashworth, and Main 2010; Janakiraman, Sismeiro, and Dutta 2009; Roehm and Tybout 2006).  From a more practical perspective, I highlight the importance of monitoring consumer sentiments not only toward a company’s own brand but also toward its competitors, and identify   4 conditions under which hatred for close competitors will be more versus less impactful. I also propose marketing interventions that can assuage consumers’ self-protection concerns, enabling brands to insulate themselves from the effects of hatred toward other close competitors.       5 Chapter 2: Theoretical Development  Hatred is a feature of human existence with widespread implications for social life. Accordingly, it has been examined from several disciplinary perspectives including social psychology, psychoanalysis, philosophy, and sociology. In general, theorizing on hatred emphasizes that it is a composite of a variety of "severely negative feelings" (Baumeister and Butz 2005, p. 87) and “unpleasant sensations” (Roseman and Steele 2018, p. 3). Hatred revolves around the perception that a hated other imposes material and psychological costs on the self that create zero-sum gains for the hated person at the expense of the hater (Gervais and Fessler 2017).  Past research has studied hatred both as an immediate emotion and a more composite feeling akin to an emotional attitude or sentiment (Halperin, Canetti, and Kimhi 2012; Opotow and McClelland 2007; Royzman, McCauley, and Rozin 2005). I adopt the latter perspective and examine hatred as an overall composite feeling toward a specific focal brand, which can be formed over relatively isolated or more extended experiences with this brand (Ben-Ze'ev 2018; Halperin et al. 2011). In line with existing research, my conceptualization of hatred explicitly incorporates the commonly identified components of anger, contempt, and disgust (Fischer et al. 2018; Roseman and Steele 2018; Sternberg 2003, 2005; Sternberg and Sternberg 2008). Two of these emotional markers (i.e., contempt and disgust) belong to the exclusion emotion family, whereas the third emotional marker (i.e., anger) belongs to the attack emotion family (Fischer and Roseman 2007). As Harris (2018) cogently points out, the complexity of hatred renders theoretical approaches developed for the study of quintessential emotions (e.g., fear, anger) less useful in studying hatred, as hatred does not always conform to such structures (e.g., unique appraisals, distinct nonverbal behaviors). Despite these complexities, it is important to examine understudied affective states such as hatred to increase understanding of how such negative   6 states shape the actions of the people experiencing them (Fischer 2018; Harris 2018; Roseman and Steele 2018).   While acknowledging that it is challenging – both conceptually and empirically (see also Gervais and Fessler 2017) – to cleanly separate hatred from its components, it is nevertheless important to discuss some of the key similarities and differences in experiencing hatred as a composite versus each of its individual components separately. First, anger by itself is characterized by a belief that it is possible to instigate change in the target of the anger (e.g., Fischer and Roseman 2007). Consequently, anger has been characterized as an inherently approach-oriented affective state (Carver and Harmon-Jones 2009). For instance, individuals might express anger during negotiations in order to change the behavior of the negotiation partner and thereby claim greater value for themselves (e.g., Sinaceur and Tiedens 2006; van Kleef, De Dreu, and Manstead 2004). Anger is also more likely to be remedied than contempt or disgust (Hutcherson and Gross 2011), the other components of hatred. Whereas a hope for a change in the target is characteristic for anger, hated others are ascribed an inherently evil and threatening character that is conceived as a stable disposition unlikely to change (e.g., Fischer et al. 2018; Rempel, Burris, and Fathi 2018). In one of the few direct empirical comparisons of anger and hatred, Halperin (2008) found support for these differences in the context of the Israeli-Palestinian conflict: whereas anger among Israeli Jews was most strongly associated with a desire for educational efforts to change and improve the out-group (i.e., Palestinians), hatred was primarily associated with desires to exclude, remove, and destroy the out-group.  Second, the experience of hatred as a composite also differs from those of contempt and disgust when experienced separately. At the heart of contempt is the perception that the other is of low social value (e.g., incompetent, cold) and unworthy of respect (e.g., Fischer and Giner-  7 Sorolla 2016; Gervais and Fessler 2017). Accordingly, the central social function of this emotion is the exclusion and disengagement of a target held in contempt (Fischer and Roseman 2007). Thus, hatred and contempt are similar to each other in terms of appraising the essence of the target as stable and unfavorable (Fischer 2011). However, the appraisals underlying contempt tend to be less damning than those of hatred as they are largely based on incompetence perceptions (Hutcherson and Gross 2011). Ample evidence suggests that competence judgments are weighted less strongly in the evaluations of others (e.g., Goodwin, Piazza, and Rozin 2014). Moreover, some targets might be hated precisely for their “competence” in successfully implementing their intentions to mistreat or exploit others (Landy, Piazza, and Goodwin 2016; Piazza et al. 2014). Akin to contempt, disgust in an interpersonal context involves a negative dispositional attribution. However, in contrast to the focus on incompetence when holding another in contempt, people experience disgust towards a target when the other is construed as morally untrustworthy based on his intentional, immoral behavior (e.g., Hutcherson and Gross 2011). While disgust by itself and hatred as a composite are both relatively resistant to change, the former differs from the latter in the greater number of potential causes and behavior leading to its generation (Hutcherson and Gross 2011). Consequently, a critical difference between contempt and disgust as separate emotional experiences on the one hand and hatred as a composite on the other hand is the absence of a motivational focus or goal to harm the target held in contempt and/or disgust, unlike that which typically occurs in hatred (Rempel et al. 2018; Sternberg 2003). The principal action tendencies vis-à-vis the focal target is to reduce contact with the target via exclusion in the case of contempt (e.g., Fischer and Roseman 2007) and moral condemnation in the case of disgust (e.g., Hutcherson and Gross 2011).   8 In consumption contexts, brand hatred might also overlap with other prevalent negative feelings, such as dissatisfaction or disappointment (e.g., Oliver 2010; Westbrook 1987) but differs from them in two important aspects: first, unlike other negative feelings, hatred is rooted in the perception of the hated brand as a self-serving, evil, and exploitative actor (Gervais and Fessler 2017). Second, whereas hatred’s central function is the elimination of the hated object from one’s environment (Fischer 2018; Roseman and Steele 2018), dissatisfaction does not necessarily preclude further engagement with the object causing the feeling and entities related to it (Oliver 2010). Third, hatred is “emotionally pluripotent” (Gervais and Fessler 2017, p. 10): such that hatred for a brand involves the experience of multiple, diverse discrete emotions depending on the behavior and well-being of the hated brand. For instance, seeing a hated brand fail in the marketplace can lead to the experience of positive emotions such as joy or schadenfreude (see also Rempel et al. 2018).  Past research in the consumer-brand context has examined how certain experiences – such as immoral brand behaviors, repeated failures, and identity violations (e.g., Grégoire, Tripp, and Legoux 2009; Romani et al. 2012; Zarantonello et al. 2016) – create negative feelings akin to hatred, and how these feelings can shape consumer behavior toward the brand. Theorists have emphasized that hatred often compels individuals to take actions against the hated entity (Rempel and Burris 2005; Rempel et al. 2018; Sternberg and Sternberg 2008). In the brand domain, some of these actions, such as complaining or negative word-of-mouth, can be instrumental and are pursued with the goal of instigating change in the hated brand (Grégoire, Laufer, and Tripp 2010). Other actions (e.g., anti-brand websites, aggressive behaviors toward employees) are purely destructive and intended to harm the brand (Johnson, Matear, and Thomson 2011; Kähr et al. 2016).    9 While much of the theorizing on hatred emphasizes these aggressive ways of engaging with the focal object, a separate stream has suggested that hatred might lead individuals to be concerned that the hated person or object poses a threat (Gervais and Fessler 2017; Royzman et al. 2005). As discussed previously, hatred frequently involves the experience of contempt and disgust, two other-distancing emotions. Supporting these notions, empirical work in the domains of marriage and the workplace suggests that hatred is more likely to be associated with avoidant rather than aggressive or destructive behaviors (Fitness 2000; Fitness and Fletcher 1993). Moreover, an fMRI study of its neural correlates suggests that hatred produces not only patterns of brain activity in areas related to anger and aggression, but also in areas related to danger and fear (Zeki and Romaya 2008). In the consumer domain, brand transgressions that generate hatred toward an offending brand can lead to the development of persistent desires to avoid the transgressing brand (Grégoire et al. 2009).   2.1 Hatred Activates Self-Protection Concerns  While the link between hatred and the rejection of the focal hated object has been generally supported, little is known about the downstream consequence of this rejection, that is, how hatred for one specific brand will impact preferences for competitors. Theorists have hinted at the possibility that hatred may impact objects beyond the focal hated entity (Opotow and McClelland 2007), but the nature of this influence in the brand domain is uncertain. One possibility suggested in prior research is that the aggressive tendencies including a desire to exact revenge on the hated brand also guides subsequent brand choices. Consequently, consumers may become “purchase activists” (Paharia, Avery, and Keinan 2014), who seek to express their sentiments and make an impact in the marketplace via their consumption choices. In the context   10 of brand hatred, consumers may become a purchase activist by choosing the closest, most direct competitor of a hated brand (Nasr Bechwati and Morrin 2003). However, as detailed below, I instead suggest that the preferences of consumers who experience hatred for competing brands is generally not guided by such aggressive motivations. Instead, their preferences will be dictated by a desire to protect the self from further harm.  Work in the context of intergroup conflict suggests that hatred increases concerns about potential danger resulting from interactions with members of the hated group (Halperin et al. 2012). Self-protection concerns are triggered when individuals feel that their physical safety or other valued resources are imperiled (Kenrick et al. 2010; Lisjak and Lee 2014). Hatred tends to arise from experiences in which others impose costs on the self, whether material (e.g., unexpected charges) or psychological (e.g., humiliation). Contemplating such experiences should make self-protection concerns salient, resulting in a focus on evading potential harm and dangers in the social and consumption environment. This focus increases vigilance to cues signaling peril and greater reliance on such cues when making decisions (Griskevicius and Kenrick 2013). When concerned about self-protection, individuals seek safety and hence, intend to make safer choices. This manifests in greater loss aversion (Li et al. 2012) and an increased propensity to engage in risk-reduction behaviors (Lisjak and Lee 2014). In the context of brand choices, I argue that brand hatred activates a general concern about self-protection, which influences brand choice even when the hated brand is not in the consideration set.  A brand’s category membership is a readily available cue that helps consumers evaluate potential consumption experiences. When consumers hate a brand from a particular subcategory (e.g., a chain hotel), they may ascribe greater potential harm to another brand that belongs to the same subcategory. The motivation to avoid danger and make safe choices should, then, shift   11 consumer preferences to brands outside the subcategory of the hated brand. Reflecting upon a hated brand increases concerns about self-protection driven by conceiving the hated brand as a threat (Royzman et al. 2005; Zeki and Romaya 2008). These concerns, in turn, shape preferences for competitors in favor of a brand from a different (vs. same) subcategory than the hated brand. Such shifts in preference are less likely to occur when consumers experience other less severe negative feelings toward brands such as the common experience of dissatisfaction. Feelings of dissatisfaction are less intensely affective in nature and are often associated with attempts to understand why certain unpleasant events have occurred (Bougie, Pieters, and Zeelenberg 2003). In contrast, the emotional experience of hatred is associated with distancing oneself to avoid further harm (Gervais and Fessler 2017; Haidt 2003; Winterich, Mittal, and Morales 2014), which in turn leads consumers to prefer safer brands from other subcategories. To summarize, formally: H1:  Hatred (vs. no hatred) for a focal brand leads to a preference for a competing brand from a different subcategory compared to a brand from the same subcategory. H2:  The effect of hatred on preferences is mediated by self-protection concerns.      12 2.2 When Hatred Leads to Preference for a Same-Subcategory Brand: An Opportunity for Revenge  Although I have argued that in general, the paramount concern for a consumer is avoiding further harm, there might be situational factors that make other concerns salient. Nasr Bechwati and Morrin (2003) demonstrated that some consumers sacrifice their own interests when presented with the opportunity to take revenge on a brand engaging in severe transgressions. Relatedly, theorists have emphasized the centrality of wanting to harm the target in the experience of hatred (e.g., Rempel and Burris 2005). Specifically, one motivational component of hatred is a desire to diminish the well-being of the hated other (Rempel et al. 2018). When a situation grants sufficient opportunity, hatred might also lead to greater endorsement of violent actions against a hated social group (Halperin 2008). Yet, in typical marketplace situations – as mirrored in the majority of the studies in my dissertation – consumers have little or no awareness of the impact of their consumption choices on a hated brand. Accordingly, as outlined previously, concerns about protecting the self are more likely to guide subsequent choices rather than a desire to harm the focal brand. However, if exposed to a cue suggesting that their choice of one particular rival would be more harmful for the focal brand than choosing a different brand (e.g., information about an extremely fierce rivalry between the same-subcategory competitors), I expect that consumers might abandon their concerns about self-protection in favor of an opportunity to hurt a hated brand. Formally: H3:  The effect of hatred on preferences is moderated by the salience of the rivalry such that when salience is high, the preference for an other-subcategory brand is attenuated.    13 Chapter 3: Overview of Studies I report nine studies, including one qualitative study, one study utilizing archival field data collected from Yelp (study 1), one correlational study (study 2), and six online experiments (studies 3-7). The qualitative study provides insights on consumer experiences that trigger brand hatred and how hatred manifests in thoughts about other brands. Next, I utilize restaurant reviews from the online portal Yelp in study 1 to test how hatred-evoking dining experiences affect consumers’ propensity to return and review other restaurants from the same subcategory as the focal brand. Study 2 uses a correlational approach to replicate the basic effect in multiple categories. I then report a series of online experiments that test my first (study 3), second (studies 4-6), and third hypothesis (study 7) in a more controlled setting. For all online experiments conducted on Amazon Mechanical Turk I used a geographic filter (excluding non-US internet protocol addresses) and included an attention check (Oppenheimer, Meyvis, and Davidenko 2009) in the first question. Only participants who answered correctly were able to participate in the studies. Participants, identified by their Amazon Mechanical Turk worker ID, were not able to participate in more than one study. For all experimental studies, unless noted otherwise, I aimed for a sample of approximately 100 participants per cell and did not exclude any participants.1 The appendix contains additional auxiliary studies as well as supplementary details and analyses for the studies reported in the main body of my dissertation.                                                    1 As a robustness check, I repeated all reported analyses for studies 3 – 7 in subsamples of participants who correctly recalled the category of the focal brand. In these analyses, I fully replicate the reported results subsequently.   14 Chapter 4: Qualitative Study Given my interest in understanding what brand hatred feels and sounds like, I worked with a market research agency to identify consumers who hated at least one brand. In selecting consumers, I sought variation in terms of sociodemographic factors and brand experiences following general tenets of purposive sampling (O’Reilly, Paper, and Marx 2012). Following Arsel (2017), I devised a flexible interview protocol designed to encourage informants to share narratives of their strong negative feelings about brands as well as their behaviors toward the focal and other competing brands. The interviews began by asking consumers to describe their experiences with brands that they came to hate. The topics emerging from the narratives provided by my respondents were further examined by semi-structured questions and probes about the emotional experience (e.g., “How did this experience you just described make you feel at the time?”) as well as how their feelings for the focal brand affected their other consumption and shopping behaviors (e.g., “How have your feelings toward the brand affected your behaviors in general?”). Using semi-structured, in-depth interviews allows for an emic perspective on the origins, content, and consequences of consumers’ hatred for brands as well as how hatred affects their subsequent perceptions, preferences, and behaviors (Arnould and Wallendorf 1994). During the interviews, I sought to capture the emergence, underpinnings, and affective composition of consumers’ stories about the brands they hate. I used idiosyncratic probing to generate narratives about informants’ brand experiences and their consequences (Kvale 1983). Within each interview I used the narratives provided by the respondents to better understand divergence and convergence of hatred for different firms. I conducted 12 interviews. Each took place at a neighborhood café of the informant’s choosing and lasted 53 minutes on average (range: 30 to 75   15 minutes). Appendix A2 contains the outline of the interview protocol and appendix A3 provides a detailed overview of my respondents and the main themes that emerged in each interview.  4.1 Data Analysis My analysis initially followed an idiographic approach in which I took notes and interpreted emerging themes (e.g., development, affective composition, and consequences of hatred) within each individual interview. Subsequently, I examined convergence and divergence across the narratives by comparing across the different in-depth interviews (i.e., a more nomothetic approach; Thompson, Locander, and Pollio 1990). During this phase, I employed axial coding procedures, involving repeated iterations between my interpretations and relevant literature about consumer-brand connections and hatred (Spiggle 1994). I subsequently organize the presentation of the qualitative date around the three central elements that arose across the different consumer narratives: how consumers develop hatred for a brand, the feelings involved in hating a brand, and finally the consequences of brand hatred. The development of brand hatred. According to the narratives of my informants, hatred for a brand emerges in a variety of industries, from retailing, to banking, to telecommunications. The different narratives suggest that hatred can develop from a one-off interaction as well as from repeated interactions unfolding over longer time spans. While there was considerable variation in the events and experiences that led my informants to develop hatred for a given brand, the stories that generated hatred for a brand tended to involve a severe transgression by the brand causing a sense of violation. Martha’s account of how she came to hate her mobile phone operator is representative of such a severe and repeated brand failure:  “I’ve been with them for 12 years and so I am a very loyal customer I feel. I don’t know the past two to three years I feel customer service has really gone downhill. You know every time you call because you have a problem, I tried to use my Fido dollars which is like a reward that a customer gets when they you know it’s a   16 percentage of their bill it’s very small. So it takes years to get anything, so I tried to get this very small travel pack from Fido and they ended up billing me for it rather than taking my credit, it took three phone calls and like three months of me still not seeing the credit on my bill to finally get it resolved. That was a very negative experience, umm and then I recently I cracked the screen on my phone and I have insurance with them and that was a nightmare, it was a third party company so they pawned me off to them so they don’t deal with it, so I dealt with them and I was like ‘You know what, I can get a new phone from somewhere else’ […]” Martha’s narrative also alludes to the notion that consumer are generally willing to forgive repeated shortcoming of brands (Joireman et al. 2013). However, Martha was more aggravated and dismayed when her efforts to resolve the issues were met with “patronizing” treatment from staff, ultimately leading her to “strongly hate Fido”. When brands and their employees act in ways that significantly reduce the expected future relationship value and signal that such violations might repeat itself (McCullough, Kurzban, and Tabak 2013), hatred is particularly likely to arise. This account is representative of a broader concern that firms do not appreciate consumers as individuals, their loyalty, or their expertise. Similar themes emerged for other consumers such as Lucy who came to a hate a local baby store. While shopping at the store several employees gave her inaccurate and misleading product information that if followed would have led her to purchase unnecessarily expensive products for her toddler. Consumers like Lucy commonly assume that some ulterior, malicious corporate motives were driving these employee behaviors (Crossley 2009; Halperin 2008). More generally, inadequate responses to customer feedback and inquiries were often one of the seeding points for brand hatred. Many narratives revolved around disinterested, discourteous, and condescending treatment from staff. Receiving such treatment led to strong negative affective reactions, and to the attribution of generalized and stable character-like traits to the hated brand (e.g., evil, faceless, immoral) and the individuals (e.g., deceitful, corrupt) associated with it (Fischer et al. 2018; Halperin 2008). Consumers were skeptical—to outright dismissive—of the willingness of these brands to change their ways, as represented by Heather, who would not contemplate returning to the hated Walmart brand unless it undertook a “fundamental character change”.    17 Oftentimes, hatred was amplified in situations where consumers felt vulnerable due to a shortage of resources necessary to successfully navigate the marketplace. Illustrative of this theme is Barry, who felt exploited by a shoe brand at a time where he was a “newcomer [to the country], [and] had very little budget”. Because Barry was facing budget constraints he was counting on the black dress shoes to last for a while. The product failure after just two months had a dramatic effect on him. Barry’s experience was exacerbated by a feeling of helplessness resulting from his limited social resources (e.g., friends who could make recommendations).    Taken together, hatred can emerge from a variety of negative experiences, and consumers are able to vividly recall such events even after time has passed. Feeling severely violated by brands (e.g., via inappropriate employee behavior) was a common theme across narratives. The experience of brand hatred. Echoing notions that hatred is an averse and unpleasant experience, my interviews revealed a certain discomfort in declaring hatred for a brand (Ben-Ze'ev 2001). Several respondents explicitly acknowledged that they were not proud or did not feel good about hating a particular brand such as exemplified in the following statement from Martha: “I mean yeah I don’t like to use the word hate but yeah I hate Fido.” A related theme was that hating a brand could prove costly in monetary and non-monetary terms, as in the case of Heather: “it was a sacrifice to stop shopping at The Gap […]. It is inconvenient because […] we have fewer options […] for buying clothes outside of standard sizes. And especially Old Navy, I think they do like all of their pants and dresses, they'll just do them plus sizes and they'll do them tall sizes and petite and so you have choices there. So, I was frustrated at losing those choices.” Respondents’ descriptions of their feelings toward the brand were characterized by negative emotions and feelings beyond disappointment or dissatisfaction. Consumer narratives were characterized by the ongoing or recalled experience of relatively “hot” negative emotions   18 such as anger and rage as well as colder negative emotions such as disgust. My respondents described being “fed up” with and “disgusted” by the behavior of the individuals managing these brands. A sense of betrayal as illustrated in the following quote from Barry can also contribute to hatred: “[I felt] like a little bit cheated on”. More generally, hatred revolves around the perception of the hated object taking advantage relative to the self because of the imposition of ongoing material and psychological costs (Gervais and Fessler 2017; Royzman et al. 2005). Echoing these notions, Mark, Tania, Anna, and Lucy all mentioned feeling intentionally neglected by the brand. Feelings of powerlessness were also frequently noted. As Heather described her sense of lack of control vis-à-vis Monsanto: “It makes me feel powerless, […] and it makes me feel regretful or sad about the future of our world”.  The narratives provided by my respondents also attest to the notion that hatred is “emotionally pluripotent” in so far that it has the potential to evoke a diverse set of discrete emotions depending on the actions and fortunes of the hated object (Gervais and Fessler 2017, p. 10). For instance, several of my respondents expressed schadenfreude when observing that hated brands were performing poorly in the marketplace or were going out of business; as represented by Heather’s comment who felt like writing a letter to The Gap and “be like aren't you guys losing money? Do you know how you could get more money?" Other consumers, however, in the face of the continued existence and success of the hated brand expressed a sense of bewilderment or irritation, as expressed by Barry, “I may be wondering why is it still in the market, because I see the store, so I guess some people still buy the stuff […]”.  Taken together, the narratives suggest that hatred for a brand comprises various discrete emotions conditioned on the actions of the hated brand. Consumers experienced "severely negative feelings" (Baumeister and Butz 2005, p. 87) toward brands they hated involving hot and   19 cold negative emotions such anger, contempt, and disgust. Subsequent exposure to the hated brand can create feelings of powerlessness or scorn depending on the relative power balance. Interestingly, most respondents remained somewhat curious about the fortunes of hated brands and reacted to failure and success of these brands with glee and bewilderment, respectively.  Consequences of brand hatred. Brand hatred led many of my respondents to minimize contact with the focal brand when shopping in the category. Several respondents noted an unpleasant sensation when they saw the logo or advertisements, and, as Tania put it when she encounters branches of the hated brand, “[I] keep walking. [I] keep walking.” When consumers could not avoid the brand due to a lack of alternatives, they changed their approach, sometimes drastically. Such a shift is exemplified in the following expression from Lucy: “So I really, now only make a point of going in if there’s something specific that I need and I know they have it in stock. […] Otherwise I would not, yeah, I wouldn’t just go in to browse.”  Consumers also actively engaged in negative word-of-mouth against the hated brand (Kähr et al. 2016). For instance, many of my respondents reported similar behaviors as exemplified in the following quote from Barry: “It was enough for me to piss me off and say bye-bye to the brand, and after I told all my friends don’t buy it. Don’t buy it!” While some consumers reported a strong desire to take actions against hated brands, consumption and purchase decisions for similar products or services were typically not the valve for such desires. As exemplified in the following quote from Heather about her motives when purchasing San Pellegrino instead of the hated Coke brand: “It is not like I am getting back at Coke [by buying San Pellegrino], it is about taking care of myself.” That is, hatred for a brand led consumers to seek brands that were distinctively different and provided a sense of safety. Such a sense of   20 safety could emerge from various properties of other brands such as offering multiple subbrands and large assortments, as was the case in the following statement from Barry: “I wanted to try something different. In a way like I said, Aldo was a small shop. […] I don’t know if it was just Aldo [products] inside, I forgot, if they have other brands, but I think they have mostly Aldo products. Like I wanted to try like maybe a big super store with many kinds of brands. Bigger choice, choice and quality, and prices […]”  In general, consumers were quick to dismiss same-subcategory rivals of a hated brand as viable alternatives or even as pawns in an effort to get back at a hated brand. As Martha put it: “I want to stick it to them. And unfortunately as a consumer all you can really do is spend your money elsewhere. I mean I can’t, I’m not going to go egg their store. As a consumer you can’t – maybe I can leave a bad review. Maybe I’ll get around to doing it. But honestly, that takes energy, I’ve already given them far more energy than they deserve […]”. Instead they looked for markers such as a category membership or quality reputation to find providers where they would feel safer when they were in the market for the products offered by the hated brand. Interestingly, this better-safe-than-sorry mentality might have adverse effects for brands that do not squarely fit into consumers’ categorization, i.e., are perceived as crossing category boundaries. The adverse impacts for such brands is well captured in the following quote from Mark about why he would not consider Nesters, a competitor of the hated brand, Loblaws:   “I feel like they have a bit of an identity crisis because they’re not really a Safeway, but they’re definitely not anything like Whole Foods or even Choices […] when I walk into Nesters I’m not really sure what they’re trying to do, or what target market, I feel like they’re trying to kind of play in the middle […].”    As can be seen in this example, consumers readily perceive category markers and use them to organize their future shopping. Hatred for a focal brand could also lead to a generally more cautious approach to life, as was the case for Barry who linked his experiences with and hatred for Aldo to his hesitation to marry to his girlfriend. Similarly, respondents routinely extrapolated from their hatred for a focal brand to a more general concern about their own well-being in the marketplace when interacting with other firms as succinctly expressed by Martha:    21 “Why do I have that fear about, specifically when it comes to cell phones, why do you have that fear that companies will wrong you in some way, you just have that assumption. I guess cause it just constantly happens in your life, you’re always dealing with something, whether it’s your bank, your internet, your car insurance, you know you’re always fighting for what you thought you were supposed to get in life.”  4.2 Discussion  My analysis of the 12 interviews revealed a number of insights regarding how hatred for a brand develops, how consumers experience brand hatred, and how it affects their subsequent behaviors. The experience of brand hatred often involves feelings of being violated and wanting to avoid further harm, which result not only in avoidance of the hated brand, but also avoidance of other brands deemed to be similar (i.e., same-subcategory competitors). For some, hatred for a brand was also characterized by a desire to cause harm and a certain pleasure at learning about the demise of the brand. This suggests that revenge and harm motivations can accompany feeling of brand hatred, but do not supersede self-protection motives in subsequent choices. Building on the insights about the emergence and effects of hatred gained from the respondents’ narratives the subsequent studies use archival and experimental methods to examine how hatred for a focal brand affects consumer preferences for competitors of this brand.        22 Chapter 5: Study 1 The goal of study 1 was to provide preliminary market-level evidence that hatred for a brand leads consumers to eschew same-subcategory competitors. For this purpose I collected data from Yelp, a popular online review site for restaurant experiences. Recent studies based on data from the same platform have predominantly focused on the reactions of readers to the reviews of other consumers (e.g., Chen and Lurie 2013; Elder et al. 2017) or the content of a specific review (e.g., Chen 2017). Instead, my first study focused on the reviewers and examined their behaviors over time by exploiting Yelp’s chronological user review history. Specifically, I treated consumers’ reviews as proxies for their consumption behavior and examined how brand hatred reflected in a review affects the reviewer’s subsequent (reviewing) behavior of same-subcategory brands. As I used data collected from Yelp, I could not directly manipulate the feelings of consumers toward brands, thus I relied on consumers’ reviews and ratings to examine how experiences that evoke hatred (i.e., a one-star review) versus indifference (i.e., a three-star review) at the same focal restaurant affect consumers’ preferences for same-subcategory competitors (as reflected by their reviewing behavior). To validate my assumption that one-star versus three-star ratings differ in the degree of negative emotions displayed, I analyzed the review text using the Linguistic Inquiry and Word Count program (LIWC; Tausczik and Pennebaker 2010). Indeed, one (vs. three) star reviews featured significantly more negative emotion words (M 1! = 2.79, SD = 2.45 vs. M 3* = 1.25, SD = 1.40, t(1090) = 13.01, p < .001), words related to hatred (e.g., contemptuous, hate, angry; M 1! = .60, SD  = 1.07 vs. M 3! = .25, SD = .58, t(1090) = 6.88, p < .001), and swear words (M 1! = .22, SD  = .59 vs. M 3! = .10, SD = .35, t(1090) = 4.25, p < .001). As some of these variables are based on the counts of words with a low hit rate (i.e., few occurrences), I also examined the proportion of one- and three-star reviews   23 that contained any hatred-related, swear, or negative emotions words. Converging with the results based on the continuous measures, I found that a significantly greater proportion of one- (vs. three-) star reviews contained hatred-related (1! = 40.2% vs. 3! = 23.1%, χ2 (1) = 36.91, p < .001), swear (1! = 20.0% vs. 3! = 10.8%, χ2 (1) = 17.93, p < .001), and negative emotions words (1! = 89.8% vs. 3! = 66.6%, χ2 (1) = 83.21, p < .001). The category marker was cuisine because a study on Amazon Mechanical Turk (N = 103; see appendix A4 for full details) documented that the type of cuisine is one of the most commonly applied criterion for restaurant choice, and, thus, a suitable categorization variable. My central prediction was that a hatred-evoking consumption episode reduces consumer desire to return to the focal subcategory, which will manifest in a lower propensity to review another restaurant offering the same type of cuisine and an increase in the duration between the focal review and the next review of another restaurant from the same subcategory.    5.1 Data I selected five popular ethnic cuisines in the United States (i.e., Chinese, Indian, Italian, Japanese, and Thai). For each cuisine, a research assistant that was supervised by me webscraped the review history of all users who gave one- and three-star reviews for the five non-chain, ethnic restaurants in San Francisco with the highest number of reviews and a maximum overall rating of 3.5. These criteria ensured a sufficient number (Mreviews = 218) of one- and three-star reviews for each restaurant. I utilized the review history of each user to compute variables capturing their prior and subsequent review activity. Notably, some consumers on Yelp review a large number of establishments per day, with some consumers writing over thirty reviews on a single day. In order to test how a specific hatred-evoking brand experience affects subsequent reviewing   24 behavior, some temporal separation between reviews is necessary. Thus, the final sample for study 1 includes 1092 Yelp users, who at most reviewed seven different establishments on any single day during the period of observation (February 1, 2005 – January 24, 2016) and wrote a one- or three-star review describing their experiences at one of the five target focal restaurants. A maximum of seven reviews per day was selected to ensure sufficient temporal proximity between the review and the actual consumption experience. In the final sample, users on average wrote 1.48 reviews per day on each day when they wrote at least one review for a total of 64,022 reviews (or 58.62 reviews per consumer).   Independent variable. The independent variable was hatred for the restaurant, which I captured in a dummy that took the value of 1 if the restaurant was rated with one-star and the value of 0 if it was rated with three-stars. There were 592 3-star reviews and 500 1-star reviews. Dependent variables. All reviews in a user’s history contain information about the restaurant’s cuisine (e.g., Italian, Indian), which I used to compute category-based variables. First, I examined whether users returned to the subcategory at any point during the period of observation (February 1, 2005 – January 24, 2016). Returning to a subcategory means that the user posts a review for another restaurant from the same subcategory as the restaurant described in the focal review (0 = No, 1 = Yes; M = .47). Second, I calculated the inter-review time (M = 975), i.e., how much time elapsed before a consumer reviewed another restaurant from the same subcategory. If a consumer did not return to the same subcategory, the inter-review time is equal to the review age computed as the date of data collection minus the date of the review. Control Variables. To better isolate the effects of the focal consumption experience on consumers’ propensity and timing of the return to the subcategory, I controlled for several review- and reviewer-specific variables (Chen and Lurie 2013). In terms of review-specific   25 controls, I accounted for review age and number of words in the review. I also included dummy variables for the subcategory of the focal restaurant. For reviewer-specific controls, I included the number of friends that a user had on Yelp and the number of reviews a user had written before and after the focal review. Appendix A5 describes the control variables in more detail and table 1 contains the descriptive statistics of all variables. Table 1. Yelp Data Descriptive Statistics   Total 3* reviews 1* reviews N 1092 592 500 Dependent Variables     Percentage category return 47% 56% 36%  Inter-review time (in days) 975 873 1096 Review-Specific Control Variables     Review age (in days) 1711.34 1778.69 1631.6  Word count 108.83 113.78 102.97 Review-Specific Control Variables     Friends 50.61 68.77 29.11  Number of reviews before 23.43 31.84 13.47  Number of reviews after 33.22 42.53 22.21  Specification. Given the nature of the two dependent measures, I used a logit model to estimate whether consumers returned to the subcategory. To estimate the time elapsed before consumers return to the subcategory I employed an ordinary least squares model.   5.2 Results The model-free evidence suggested that the percentage of consumers who returned to the subcategory (e.g., an Italian restaurant if the focal hated restaurant was Italian) was almost 20% lower among those who wrote a hateful versus those who wrote an indifferent review (returned 1* = 36.2% vs. returned 3* = 55.7%, χ2(1) = 41.58, p < .001, d = .40) and 223 days more passed between the focal review and the return to the subcategory (M 1* = 1096.21, SD = 965.22 vs. M 3*   26 = 873.01, SD = 971.47, t(1090) = 3.79, p < .001, d = .23). To more formally test H1 (see table 2), binary logistic regressions with controls produced a significant negative effect of the hatred dummy (βexp = -.33, SE = .15, Wald χ2(1) = 4.49, p = .034) on subcategory return. In addition, the hatred dummy significantly increased the inter-review time (b = 159.18, SE = 52.16, t(1081) = 3.05, p < .01). Table 2. Propensity and Timing of Consumers’ Subcategory Return  Subcategory Return (Binary) Inter-Review Time (Continuous) Variable Coefficient SE Coefficient SE  Hatred (1!=1; 3!=0) -0.3281 ** 0.1547 159.18 *** 52.16	       Review-Specific Control Variables    Review age 0.0002 ** 0.0001 0.49 *** 0.03	  Word count -0.0006 0.001 -0.03 0.31	  Reviewer-Specific Control Variables    Friends 0.0003 0.0015 0.34 ** 0.14	  Number of reviews before -0.0026 0.0025 0.53 0.66	  Number of reviews after 0.0535 *** 0.005 -5.15 *** 0.47	  Restaurant Dummies      Indian 0.5655 ** 0.2339 -76.87 79.69	   Italian 0.8999 *** 0.2068 -271.40 *** 70.45	   Japanese 1.0134 *** 0.2257 -331.47 *** 74.77	   Thai -0.4867 * 0.258 159.86 ** 79.83	       Constant -1.6866 *** 0.2325 303.05 *** 75.26      N = 1092     Pseudo R2 .265    Adjusted R2    .316  * p < .10. | ** p < .05. | *** p < .001.   Additional analyses and robustness checks. To ensure that the effects were driven by feelings about a particular brand (e.g., a specific Indian restaurant) rather than hatred of the whole subcategory, I used three different ways to account for reviewers’ category-level preferences: first, I incorporated a dummy capturing whether this was the first time a user reviewed a particular subcategory and interacted this with the hatred dummy. Second, I controlled for same-subcategory visits as a share of total reviews prior to the focal review. Third,   27 I also accounted for the prior valence of same-subcategory reviews with dummies. Across all three models, the effect of the hatred dummy was robust to adding these controls and there was no consistent evidence for the effect of prior subcategory experiences and preferences. Appendix A6 contains full details of these and all other robustness checks discussed in this section.  A second set of robustness checks focused on the content of the reviews. In one model I used the net negativity of the review as a continuous independent variable (i.e., negative minus positive emotion words; Chen 2017) rather than the star dummies and replicated the patterns found in my main analyses. I also used these net negativity scores to identify the most negative reviews given that these reviewers are most likely to have experienced hatred. Following prior research (Berger and Milkman 2012), two independent coders coded a subset of reviews (N = 220, 89% 1* reviews) for how reviewers felt about the service (1 = rude; 7 = excellent) and the food (1 = horrible; 7 = excellent). Interrater reliability was acceptable for both variables (Krippendorff’s α = .77 and α = .71 for service and food, respectively). I averaged the coders’ ratings and used these variables to predict subcategory return and inter-review time: in both models only service perceptions predicted subcategory return (βexp = .37, SE = .16, Wald χ2(1) = 5.57, p = .018) and inter-review time (b = -116.01, SE = 45.12, t(208) = 2.57, p = .011). The effect of food perceptions was small and not significant in either model (both ps >.58). To summarize the robustness checks discussed in this section, appendix A7 contains a model multiverse analysis (Steegen et al. 2016) that provides an overview of whether and how much the conclusions from my analysis of the Yelp data change depending on the different specifications. The multiverse analysis suggests that my central result is relatively robust to different specifications.      28 5.3 Discussion The results of study 1 suggest that brand hatred shifts consumer preferences by reducing the probability that consumers will return to the same subcategory and increasing the time between the focal experience and the return to the subcategory. Importantly, these results are unlikely to be driven by category hatred rather than hatred for a specific restaurant for three reasons: First, an analysis of a subset of coded particularly negative reviews suggested that the experiences at a particular restaurant (i.e., the service they received) rather than perceptions of the cuisine were driving reviewing behavior after the focal experience. Second, one of the raters also coded whether the reviews contained any generalized positive or negative sentiments about the subcategory. This was the case in only 10% of all reviews, and excluding these cases does not change the pattern or significance of the supplemental analyses reported above. Finally, accounting for category-level preferences by incorporating the users’ review history on Yelp does not alter the main conclusions from study 1. Taken together, this suggests that hatred for a specific restaurant, rather than category hatred, drives the effect.      29 Chapter 6: Study 2 The goal of study 2 was to examine whether hatred produces similar effects in a variety of categories for this purpose I asked consumers whether they hated four well-known brands that are known to polarize consumer feelings – Ryanair, Starbucks, BP, and Volkswagen (Gander 2015; Luo, Wiles, and Raithel 2013). Subsequently, I examined how consumers’ feeling of hatred for one brand affected their preferences for same- versus other-subcategory competing brands in each of the industries, i.e., airlines, coffee shops, gas stations, and car rental.   6.1 Method Participants and design. Using Prolific Academic, an online subject pool (Peer et al. 2017), I conducted a survey with 198 UK consumers (60% female, Mage = 35.2). After being presented with a list of brands, participants indicated how they felt about each of the four brands by dragging each of the brand names either into a box with the heading "I hate this brand" or another one with the title "I do not hate this brand". Consumers’ allocations of the brand names were used to create a hatred dummy for each of the four focal brands (0 = no hatred, 1 = hatred). Next, I captured consumer preferences for competing brands by asking subjects to imagine four consumption scenarios, one for each industry, in which they indicated whether they would prefer a brand from the same subcategory (e.g., in the case of Ryanair, another low-cost airline, i.e., easyJet) or a brand from another subcategory (e.g., British Airways, a full-service airline; see appendix A8 for all four scenarios). For instance, the brief scenario in the airline setting read, "Imagine you were about to book a flight. While searching for flights, you find that both British Airways and easyJet offer a flight to your destination. Please indicate which of these two airlines you would prefer". The order of the four consumption scenarios was counterbalanced. For each   30 scenario, participants responded to a single item nine-point scale anchored at 1 (definitely British Airways [or respective other-subcategory brand]) and 9 (definitely easyJet [or respective same-subcategory brand]). I also controlled for age, gender, nationality, and family income. As all results remain virtually unaffected when demographic controls are entered, they are not discussed any further.   6.2 Results  In three of the four different categories 25% or more of all consumers reported feelings of hatred for the focal brand (i.e., Ryanair, 29%; Starbucks, 26%; or BP, 25%). In contrast, only 7% reported hatred for Volkswagen. To test my hypothesis that hatred for a focal brand leads to a preference shift toward the brand from the other category, I ran separate regressions with each of the hatred dummies as the independent variable and consumers’ relative preferences for competing brands as the dependent variable. I found evidence consistent with H1 in three of the four categories: specifically, hatred (vs. non-hatred) for a brand led to a reduced preferences for same-subcategory competitors for airlines (b = -1.37, SE = .36, t(196) = -3.42, p = .001), coffee shops (b  = -1.26, SE = .46, t(196) = -2.73, p = .007), and gas stations (b = -.71, SE = .37, t(196) = 1.92, p = .056). In the fourth category, car rentals, I did not find that hatred for Volkswagen led consumers to prefer Lexus to Mercedes. Interestingly, in this context I found a non-significant trend in the opposite direction (b = 1.06, SE = .65, t(196) = 1.64, p = .104). This trend should be interpreted with caution as only 7% of the sample expressed hatred for Volkswagen.   6.3 Discussion Study 2 replicates the patterns of study 1 using a correlational approach in multiple categories. Specifically, hatred (vs. non-hatred) leads to a preference shift away from   31 competitors in the same subcategory toward competitors in a different subcategory. Consumers who hate a specific brand do not prioritize such same-subcategory competitors to more distant competitors from another subcategory; in fact, they seem to avoid them.       32 Chapter 7: Study 3  Studies 1 and 2 provided initial evidence that brand hatred reduced the tendency to return to the same subcategory. Admittedly, several limitations might arise from using archival and correlational data. First, the one-star reviews used in study 1 were characterized by more negative feelings and words related to hate, I cannot definitively conclude that every one-star review involved hatred-evoking experiences. Similarly, I cannot reliably test my assumption that a three-star review represented a neutral or indifferent stance. Further, an exploratory examination of the review texts used in study 1 suggests that the feelings I examine were largely formed based on a single experience rather than multiple interactions. With respect to study 2, it is not clear what experiences underlie consumers’ feelings of hatred for the different brands. Study 3 addressed these concerns by experimentally manipulating consumers’ feelings using a scenario approach in a context characterized by more extended interactions, taxation advisory services. In addition, I compared three feelings consumers may experience toward providers: hatred, indifference, and dissatisfaction. I used a factorial design with three levels (feeling: hatred vs. dissatisfaction vs. indifference) to test H1, that hatred, but not dissatisfaction or indifference, produces a preference for a provider from a different subcategory.  7.1 Method Participants and design. Two hundred and ninety-six participants (55% female, Mage = 36.4) were recruited online from Amazon Mechanical Turk. Participants read a scenario about an experience they had with their advisor, George Adams, during the last tax season. I varied their experiences such that it made them feel indifference, dissatisfaction, or hatred for their tax advisor. Building on the insights gained from the qualitative study, the provider in the hatred condition did not only provide insufficient service, but also acted in unscrupulous and malicious   33 ways (the full scenario is included in appendix A9). For instance, in the hatred condition the provider acted brazenly, was unapologetic after making errors, and overcharged the participant.   As it was critical to ensure that my manipulation induced similar levels of dissatisfaction in the dissatisfaction and hatred conditions, while simultaneously inducing stronger hatred-related emotions (i.e., anger, contempt, and disgust) in the hatred condition, I conducted a pretest with a separate group of Amazon Mechanical Turk workers (N = 94, 45.7% female, Mage = 36.2). Participants read the same scenario as in the main study and, then rated the extent to which this experience would lead them to feel a host of 17 negative emotions ( 1 = "not at all" to 7 = "very much"; Romani et al. 2012; for full details on the pretest see appendix A10). A factor analysis produced three distinct factors: hatred (e.g., disgusted, scornful; α = .97), dissatisfaction (e.g., dissatisfied, discontented; α = .94), and more general negative affect (e.g., unhappy, upset; α = .90). A one-way ANOVA showed significant differences between conditions (F(2, 91) = 41.82, p < .001) on the hatred-related emotions. A post-hoc test using Tukey’s correction showed that participants experienced higher levels of hatred-related emotions in the hatred (M = 5.41, SD = 1.37) than in the dissatisfaction condition (M = 3.59, SD = 1.37, p < .001). In contrast, there were no significant differences between the hatred and dissatisfaction conditions on the emotions related to dissatisfaction (Mhatred = 6.09, SD  = 1.41 vs. Mdissatisfaction = 5.76, SD  = .97, p > .58) or general negative affect (Mhatred = 4.39, SD  = 1.29 vs. Mdissatisfaction = 4.01, SD = .99, p > .45), suggesting that the manipulation was effective. As expected, participants in the hatred and dissatisfaction conditions experienced greater hatred, dissatisfaction, and general negative affect than participants in the indifference condition (all ps <. 001). In the main study, after reading about their experiences with George Adams, participants were told that they had identified four potential service providers that could help them with filing   34 their taxes this year. Two of these options were from the same subcategory (i.e., individual tax consultants), whereas the other two options were providers from a different subcategory, namely the local branches of national tax filing firms (i.e., H&R Block and Liberty Tax Service). I summed the same- and other-subcategory choices such that my dependent variable captures whether participants choose a provider from the same subcategory or not. As a robustness check I also analyzed the four options separately, and found that the results hold. Since it aligns closer with my conceptualization I only report the analysis of the summed choices. After participants made their choice, I asked them to briefly describe the rationale for their choice. Finally, participants completed fourteen items adapted from the Triangular Hatred Scale (THS; Sternberg and Sternberg 2008; see appendix A11 for the full scale) as a manipulation check for the feelings manipulation and the recall question regarding the subcategory of the focal provider.  7.2 Results Feeling induction manipulation checks. To assess the effectiveness of the feeling manipulation I conducted a two-way ANOVA on the mean of the fourteen THS manipulation check items (α = .98), which produced a significant effect of condition (F(2, 293) = 222.06, p < .001). Follow-up post hoc tests using Tukey’s correction showed that participants in the hatred condition experienced significantly more hatred for their last year’s tax advisor George Adams (M = 5.41, SD = 1.37) than those in the dissatisfaction (M = 3.61, SD = 1.67) or indifference condition (M = 2.20, SD = 1.59, both ps < .001). Participants in the dissatisfaction condition experienced greater levels of hatred than in the indifference condition (p < .001). Almost all participants (96%) correctly recalled that George Adams was an individual tax consultant.    35 Choice of tax advisor. The choice share of the same-subcategory options differed across the three different feelings conditions (χ2(1) = 22.98, p < .001, d = .58). Specifically, participants in the hatred condition displayed a smaller propensity to choose providers from the same subcategory (23.2%) than those in the dissatisfaction (50.5%) or indifference condition (53.9%). The choice share of providers from the same subcategory was smaller than chance in the hatred condition (χ2(1) = 28.37, p < .001, d = 1.27), but not in the dissatisfaction (χ2(1) = .01, p > .91, d = .02) or indifference condition (χ2(1) = .63, p > .42, d = .16). In a first binary logistic regression with provider choice as the dependent variable (0 = other-subcategory, 1 = same-subcategory) I used two reverse-coded dummies (-1 = dissatisfaction/indifference, 0 = hatred) with hatred serving as the baseline. The effects of the dissatisfaction (βexp = -1.22, SE = .31, Wald χ2(1) = 14.98, p < .001) and the indifference dummy (βexp = -1.35, SE = .31, Wald χ2(1) = 19.03, p < .001) were negative and significant suggesting that participants in the hatred condition were significantly less likely to choose a provider from a same subcategory. In a separate binary logistic regression using indifference as the baseline, I found that there was no significant difference between the dissatisfaction and the indifference condition (βexp = -.14, SE = .28, Wald χ2(1) = .23, p > .63).   7.3 Discussion Study 2 experimentally replicated the findings from study 1 and 2 in a controlled setting that eliminates concerns inherent to the Yelp and correlational data. The results also provide support for the claim that brand hatred is distinct in its ability to shift consumer preferences. Only when consumers experienced hatred for a provider, but not when they were indifferent or dissatisfied, were they significantly more likely to prefer a provider from a different subcategory.    36 Chapter 8: Study 4a The main goal of study 4a was to shed light on the mechanism that I propose underlies my effect, namely self-protection concerns (H2). A potential concern with study 3 is that the other-subcategory was the national brands (vs. individual providers). Griskevicius and Kenrick (2013) suggest that when self-protection concerns guide behavior consumers might prefer more established brands. To address this concern, I manipulated the category of the focal brand in study 4a to demonstrate that the hatred-induced preference shifts were not merely a function of choosing more established service providers. To assess the mediating role of consumers’ self-protection concerns study 4a used a 2 (feelings toward the focal brand: indifference vs. hatred; between-subjects) x 2 (category: independent vs. chain; between-subjects) x 2 (purchase likelihood: independent vs. chain; within-subject) mixed-design.   8.1 Method Participants and design. Six hundred and two participants (60% female, Mage = 35.9), recruited from Amazon Mechanical Turk, read a scenario about a consumption experience at a hotel called the Waterside Inn. In many service categories, a salient difference between brands that consumers might use to form categories is whether a brand belongs to a chain (e.g., Hilton) or is independently owned and run. Independent as well as chain brands have offerings that are comparable, if not identical, with respect to their services, positioning, and pricing. Thus, I manipulated whether the Waterside Inn was a chain or an independent hotel. A brief pretest with 100 Amazon Mechanical Turk workers (58.0% female, Mage = 33.0) validated this category marker. Participants were presented with a list of four hotels (two independent and two chain hotels, see appendix A12) and were asked to drag-and-drop the hotels into categories.   37 Consumers were significantly more likely than chance to place chain hotels together and independent hotels together (90%; χ2(1) = 64.00, p < .001), supporting using hotel type as the category marker.  In the main study, feelings were manipulated by a description of their stay at the Waterside Inn. In the hatred condition the stay was described as “incredibly horrible” with experiences including rude staff, a dirty room, and overcharging. In the indifference condition the stay was described as “in line with expectations” with pleasant staff, a clean room, and reasonable charges (see appendix A13 for full details). After reading about their experiences, participants were told that they were about to return to this city but were not going back to the Waterside Inn. Next, I measured out mediator – self-protection concerns – by asking participants to imagine that while looking at potential options for their stay they were reflecting on their experiences at the Waterside Inn. Participants responded to three items measuring self-protection concerns ("At this moment, protecting my needs is at the center of my focus", “At this moment, I am more focused on protecting myself”, and “At this moment, I am concerned about protecting my own needs and interests”; Winterich et al. 2014).  Next, participants were shown a booking portal screenshot providing information about two other hotels (see appendix A14). The brand names indicated subcategory of each hotel. Depending on the category of the focal brand (i.e., chain or independent), the same and other subcategory was either the chain or the independent hotel. For my dependent measure participants reported purchase likelihood for both hotels on a nine-point scale (1 = "not at all likely", 9 = "very likely"; Sundie et al. 2011). Finally, participants completed the same THS manipulation check items as in study 2 as well as a recall question (appendix A15) to assess the effectiveness of my manipulation of the category of the focal brand and demographics.    38 8.2 Results  Manipulation check. To assess the effectiveness of the category manipulation, I used a binary logistic regression with the recall of the hotel brand as a chain (0 = no, 1 = yes) as the dependent variable and category, hatred, and their interaction as the predictors. This analysis only produced a significant main effect of the category dummy (βexp = 5.26, SE = .52, Wald χ2(1) = 103.59, p < .001). No other effects were significant (all ps > .33). To further verify my manipulation, I examined the correct recall across conditions (ranging from 79.3% to 86.3%), and ran a separate logistic regression with correct recall as the dependent variable and category, hatred, and their interaction as predictors that did not produce a significant effect (all ps > .24). Thus, the manipulation of the category of the focal brand was successful and did not affect participants’ propensity to recall the category of the focal brand. With respect to the feeling manipulation, a two-way ANOVA on the mean of the fourteen THS manipulation check items (α = .98) with feelings toward the focal brand and its category as between-subjects factors produced only a main effect for feelings toward the focal brand, such that participants in the hatred conditions experienced more hatred toward the Waterside Inn than those in the indifference conditions (Mhatred = 6.30, SD  = 1.51 vs. Mindifference = 1.86, SD  = 1.27, F(1, 598) = 1509.59, p < .001). Neither the effect of category (F(1, 598) = 1.41, p > .23) nor the interaction between feelings and category (F < 1, p > .85) were significant. Purchase likelihood. I used a mixed model ANOVA to assess how brand hatred (vs. indifference) and category of the focal brand (i.e., chain vs. independent) affected the purchase likelihood for the same- versus other-subcategory competitor. This analysis produced a significant three-way interaction between feelings, category, and purchase likelihood (F(1, 598) = 61.85, p < .001, ηp2 = .094, see figure 1). The two-way interaction between purchase likelihood   39 and the category of the focal hotel was also significant (F(1, 598) = 61.33, p < .001, ηp2 = .093) as was the main effect for purchase likelihood (F(1, 598) = 20.39, p < .001, ηp2 = .033). The two-way interaction between feelings and purchase likelihood was not significant (F < 1, p > .72). In line with H1, hatred induced a preference for the brand from the other subcategory, both when the focal hotel was a chain (Mindependent = 6.86, SD = 1.65 vs. Mchain = 5.62, SD  = 2.17, F(1, 598) = 32.12, p < .001, ηp2 = .051; figure 1, panel A) and an independent hotel (Mchain = 7.17, SD  = 1.67 vs. Mindependent = 4.81, SD = 2.25, F(1, 598) = 99.54, p < .001, ηp2 = .143; figure 1, panel B). Unexpectedly, I found when their initial stay made them indifferent, participants staying at a chain (Mchain = 6.63, SD  = 1.70 vs. Mindependent = 6.15, SD = 1.92, F(1, 598) = 4.09, p = .044, ηp2 = .007; figure 1, panel A) and an independent hotel (Mchain = 7.04, SD  = 1.43 vs. Mindependent = 6.57, SD = 1.56, F(1, 598) = 4.51, p = .034, ηp2 = .007; figure 1, panel B) exhibited a preference for the chain hotel. I return to this finding in the discussion of the study.     40 Figure 1. Purchase Likelihood for Chain and Independent Hotels (Study 4a) Panel A: Focal Brand = Chain Hotel   Panel B: Focal Brand = Independent Hotel Note: The error bars represent 95% confidence intervals. Self-protection concerns. To examine the underlying role of self-protection concerns, I first tested whether hatred led to greater self-protection concerns. A two-way ANOVA on the self-protection measure (α = .90) with feelings and category as between-subjects factors produced only a main effect for feelings toward the focal brand, such that participants in the hatred conditions (M = 6.29, SD = .84) exhibited greater self-protection concerns than participants in the indifference conditions (M = 5.35, SD = 1.24, F(1, 598) = 116.39, p < .001, ηp2 = .163). No other effects were significant (all Fs < 1, all ps > .42).   41 Subsequently, I conducted a moderated mediation analysis in which feelings toward the focal brand acted as the independent variable (0 = indifference, 1 = hatred), category of the focal brand as the moderator (0 = independent, 1 = chain), and generalized self-protections concerns as the mediator (see figure 2). As the mediation analyses are regression-based, I used the difference score (chain – independent) as the dependent variable (see Kumar 2005). Thus, positive (negative) values on this variable indicate a preference for the chain (independent) hotel. I tested whether hatred increases self-protection concerns and whether the effects of self-protection concerns on the preference for the chain is moderated by the category of the focal brand (PROCESS model 14; Hayes 2018).  Figure 2. Moderated Mediation Model (Study 4a)  The index of moderated mediation was significant, as the confidence interval from 5,000 bootstrap samples does not include zero (-.86, SE = .30, 95% CI: - 1.33 to -.44) implying that the conditional indirect effects differ significantly depending on the category of the focal brand. Hatred (vs. indifference) predicted self-protection concerns (b = .94, SE = .09, t(600) = 10.82, p < .001). The interaction of self-protection concerns and the category of the focal brand predicted preferences (b = -.92, SE = .21, t(597) = 4.46, p < .001). In addition, the main effect of self-protection was significant (b = .46, SE = .15, t(597) = 3.12, p < .01) as was the main effect of the moderator (b = 3.52, SE = 1.22, t(597) = 2.89, p < .01). The main effect of hatred was not   42 significant (p > .87). Next, I examined the critical indirect paths. When consumers stayed at the chain hotel, the conditional indirect effect of hatred via self-protection concerns was negative and significant (a × b = -.43, SE = .18, 95% CI: -.80 to -.09). In contrast, if the focal brand was an independent hotel, the conditional indirect effect of hatred through self-protection concerns was positive and significant (a × b = .43, SE = .13, 95% CI: .20 to .69).  8.3 Discussion  Consistent with H2, study 4a demonstrated that brand hatred increased consumers’ general concerns about self-protection. These concerns, in turn, led consumers to prefer a brand from the other subcategory than the focal brand that evoked hatred. Unexpectedly, I also found that consumers displayed a general preference for the chain brand and a main effect of self-protection concerns leading to a greater preference for the chain brand. While unexpected, this finding dovetails with theorizing that concerns about self-protection render more established, national brands as more attractive for consumers (Griskevicius and Kenrick 2013). Importantly, I found a preference reversal when the focal hotel was a chain and a significant amplification of consumers’ preferences for other-subcategory brands when the focal hotel was an independent.      43 Chapter 9: Study 4b The main goal of study 4b was to corroborate the central role of self-protection concerns in driving preferences when consumers hate a focal brand (H2). Past research, has suggested that a desire for revenge and vengeance might also influence consumers’ subsequent choices of competing brands when consumers hate a focal brand (e.g., Nasr Bechwati and Morrin 2003; Nasr Bechwati and Morrin 2007). Accordingly, I also measured consumers’ desire for revenge against the focal brand as a competing mediator in study 4b. To further probe the robustness of self-protection concerns in determining subsequent consumption, I also manipulated whether consumers’ preferences would be visible to the focal brand or not. Prior work on revenge has suggested that avengers are particularly satisfied if they are able to send a message to offenders (Gollwitzer and Denzler 2009) as would be the case if the offending brand is aware of the subsequent choice of a (close) competing brand. To summarize, study 4b used a 2 (feelings toward the focal brand: indifference vs. hatred; between-subjects) x 2 (observability of subsequent choice to the focal brand: not visible vs. visible; between-subjects) x 2 (purchase likelihood: independent vs. chain; within-subject) mixed-design. While I used the same hotel paradigm as in study 4a, the initial stay of all participants was at an independent Waterside Inn hotel.  9.1 Method Participants and design. Four hundred and five (49% female, Mage = 35.3), recruited from Amazon Mechanical Turk, read a scenario about a consumption experience at a hotel called the Waterside Inn. Feelings were manipulated by a description of their stay at the Waterside Inn in the same way as in study 4a. Subsequently, I manipulated the observability of the participants’   44 choice to the focal brand. In the non-visible condition that constitutes an exact replication of study 4a, participants were first asked to rate their satisfaction with their stay at the Waterside Inn and then were shown the booking portal screenshot used in study 4a. In contrast, in the visible condition, participants learned that on the final day of their stay they found a questionnaire from the hotel on their desk, which included (on two separate pages) the same questions as in the non-visible condition (i.e., satisfaction and choices for competing brands). Both were presented under the guise that the hotel was interested in better understanding the views of its customers. After indicating their preferences, participants responded to the same three items tapping into their generalized concerns about self-protection (α = .92) and five additional items adopted from Grégoire et al. (2010) measuring their motivation for revenge against the Waterside Inn (e.g., “I want (or wanted) to take actions to get the Waterside Inn in trouble”, “I want (or wanted) to punish the Waterside Inn in some way”; α = .97). Finally, participants completed the same THS manipulation check items and provided their demographic information.   9.2 Results  Manipulation check. To assess the effectiveness of the feeling manipulation, a two-way ANOVA on the mean of the fourteen THS manipulation check items (α = .98) with feelings toward the focal brand and its choice observability factors produced only a main effect for feelings toward the focal brand, such that participants in the hatred conditions experienced more hatred toward the Waterside Inn than those in the indifference conditions (Mhatred = 6.11, SD  = 1.54 vs. Mindifference = 1.80, SD  = 1.53, F(1, 401) = 793.89, p < .001). Neither the effect of choice   45 observability (F(1, 401) = 2.66, p > .10) nor the interaction between feelings and category (F < 1, p > .44) were significant. Purchase likelihood. A mixed model ANOVA assessed how feelings (hatred vs. indifference) and observability of the choice (not visible vs. visible) affected the purchase likelihood for the same- versus other-subcategory competitor. This analysis produced a significant two-way interaction between feelings and purchase likelihood (F(1, 401) = 17.49, p < .001, ηp2 = .042). This interaction qualified the main effect for purchase likelihood (F(1, 401) = 36.70, p < .001, ηp2 = .084). No other effects were significant (all Fs < 1, all ps > .60). In line with H1, hatred induced a preference for the brand from the other subcategory regardless of whether the choice was visible to the focal brand (Mother-subcategory = 6.90, SD = 2.14 vs. Msame-subcategory = 5.48, SD  = 2.42, F(1, 401) = 22.75, p < .001, ηp2 = .054) or not (Mother-subcategory = 7.29, SD = 1.64 vs. Msame-subcategory = 5.57, SD  = 2.52, F(1, 401) = 31.00, p < .001, ηp2 = .072). In contrast, when consumers were indifferent they did not exhibit a significant preference for one of the two brands regardless of whether their consumption choice was observable for the focal brand (Mother-subcategory = 6.69, SD = 1.76 vs. Msame-subcategory = 6.40, SD  = 1.70, F < 1, p > .35) or not (Mother-subcategory = 6.38, SD = 1.65 vs. Msame-subcategory = 6.10, SD  = 1.80, F < 1, p > .35). Self-protection concerns and desire for revenge. To examine which motivations underlie the preference shifts induced by brand hatred, I first tested whether hatred increased concerns about self-protection and desire for revenge. A two-way ANOVA on the self-protection measure with feelings and choice observability as between-subjects factors produced only a main effect for feelings toward the focal brand, such that participants in the hatred conditions (M = 6.00, SD = 1.11) exhibited greater self-protection concerns than participants in the indifference conditions   46 (M = 5.32, SD = 1.28, F(1, 401) = 32.50, p < .001, ηp2 = .075). Neither the effect of observability (F < 1, p > .41) nor the interaction were significant (F(1, 401) = 2.81, p > .09). A similar two-way ANOVA on the desire for revenge measure also only produced a main effect of feelings such that consumers had a greater desire for revenge when their stay made them hate the Waterside Inn (M = 4.37, SD = 1.88) than when it did not (M = 1.49, SD = 1.08, F(1, 401) = 352.00, p < .001, ηp2 = .467). No other effects were significant  (all Fs < 1, all ps > .34). Subsequently, I conducted a competitive mediation analysis in which feelings toward the focal brand acted as the independent variable (0 = indifference, 1 = hatred) and generalized self-protections concerns and desire for revenge as the mediators. I also controlled for the observability of the choice (0 = not visible, 1= visible). Mirroring study 4a, the dependent preference measure was the difference score, which I obtained by subtracting the preference for independent hotel from the chain hotel preference. As all participants imagined staying at an independent hotel, higher values on the dependent variable indicate a higher preference for the brand from the different subcategory. I tested whether hatred increased self-protection concerns and desires for revenge and whether these in turn mediated the effect of hatred on preferences (PROCESS model 4; Hayes 2018). Hatred increased consumers’ generalized concerns about self-protection (b = .68, SE = .12, t(402) = 5.68, p < .001) and their desires for revenge (b = 2.88, SE = .15, t(402) = 18.79, p < .001). The observability of the choice was not a significant predictor of self-protection concerns (b = .09, SE = .12, t(402) = .77, p > .44) and desires for revenge (b = .15, SE = .15, t(402) = .95, p > .34). While hatred predicted both self-protection and desires for revenge, only self-protections concerns (b = .29, SE = .13, t(400) = 2.26, p = .02), but not desire for revenge (b = -.02, SE = .10, t(400) = .81, p > .81), was a significant predictor of preferences. The main effect of hatred was significant as well (b = .58, SE = .21, t(400) = 2.73, p   47 < .01), but the main effect of choice observability was not (b = -.18, SE = .31, t(400) = .58, p > .56). Self-protection concerns mediated the effect of hatred on preferences as the indirect effect of hatred via self-protection was significant with the bias corrected bootstrap 95% confidence intervals based on 5,000 bootstrap samples that did not include zero (.20, SE = .09, 95% CI: .03 to .40). In contrast, the indirect effect of hatred via desire for revenge was not significant (-.07, SE = .33, 95% CI: -.72 to .59).  9.3 Discussion  Study 4b further substantiated the critical role of concerns about self-protection in driving choices after hatred-evoking consumption experiences. In addition to offering a direct replication of study 4a, this study also examined the role of a competing mediator identified in prior research – desire for revenge. While I replicate prior research in so far that hatred led to an increased desire for revenge, subsequent mediational analyses revealed that only self-protection concerns, but not desires for revenge, mediated the effect of hatred on subsequent preferences. The results of study 4b also indicate that merely making the choice of competing brands observable to the focal brand is not to sufficient to relegate consumers’ concerns about self-protection and motivate them to choose a closer competitor of the hated brand.       48 Chapter 10: Study 5 The goal of study 5 was to provide further evidence for my self-protection account by using a moderation-of-process approach (Spencer, Zanna, and Fong 2005) that directly manipulates the extent to which the same-subcategory brand addresses consumers’ concerns about self-protection. As in studies 4a and 4b, the quality of the options was equivalent, but I manipulated the extent to which each option would assuage concerns about self-protection. Specifically, I manipulated the variance of customer ratings for the same- and other-subcategory provider. Many review portals (e.g., TripAdvisor and Yelp) display the distribution of consumer ratings next to the average ratings. Whereas existing research documents that distributional properties such as variability in consumer opinions and ratings might influence consumer preferences (e.g., He and Bond 2015; Rozenkrants, Wheeler, and Shiv 2017), I suggest (and validate in a pretest) that greater variance (i.e., more polarized ratings) in consumer ratings prompts the impression that an offering is relatively less safe as it increases the risk of a highly undesirable service experience (Wang, Liu, and Fang 2015). Accordingly, an offering with greater variance in consumer ratings is less capable of soothing concerns over self-protection. I thus propose that when a brand from the other subcategory has greater variance in its ratings (i.e., more extremely negative and positive ratings) consumers’ preferences reverse to prefer a same-subcategory brand with less variance in ratings. To test this, I used a 2 (feelings toward the focal brand: indifference vs. hatred; between-subjects) x 2 (variance in the ratings of the other-subcategory brand: equal vs. higher; between-subjects) x 2 (purchase likelihood: independent vs. chain; within-subject) mixed-design. In the equal variance condition, I expect to replicate the effects documented in the previous studies, whereas in in the higher variance condition, where   49 the other-category option does appear to be not well suited to address consumers’ concerns about self-protection, the preferences should be reversed.   10.1 Method Participants and design. Three hundred and ninety-five participants (51% female, Mage = 34.9), recruited from Amazon Mechanical Turk, read a scenario about a consumption experience at a hotel similar to the previous two studies. All participants imagined staying at the Waterside Inn, which was branded as an independent hotel. After inducing hatred or indifference toward the Waterside Inn, participants were told that they were about to return to this city. Next, I used a modified booking portal screenshot modeled after TripAdvisor displaying the ratings as well as distributions of the ratings for the same- and the other-subcategory brand (see figure 3).  The modal rating in all conditions for all brands was “very good”, 4/5 stars (48% of all reviews) and average rating was kept constant at 3.5/5. In the equal variance condition, which serves as a conceptual replication, both hotels had an identical, unimodal review distribution with the overwhelming majority of remaining reviews being “average”, 3/5 (47% of all reviews). In contrast, in the higher variance condition the other-subcategory brand had a much more polarized rating distribution. This brand had a similar share of “excellent” (26%) and “terrible” (24%) ratings. Using these distributions, a pretest (N = 100; Amazon Mechanical Turk) confirmed that consumers indeed perceive a company with lower variance in its ratings (M = 5.32, SD = 1.41) as better suited to address concerns about self-protection than a brand with more variability (M = 4.19, SD = 1.72, t(99) = 4.16, p < .001, d = .50; “Hotel A/B seems like a safe choice to me.”; 1 = Strongly Disagree to 7 = Strongly Agree).       50 Figure 3. Variance Manipulation (Study 5) Choice set in the higher variance in the ratings of the other-subcategory brand (treatment) 10.1.1.1.1.1.1 Choice set in the equal variance in the ratings of the other-subcategory brand (replication)      In the main study, participants reported their purchase likelihood for each of the brands on the same nine-point scales as in studies 4a and 4b. Finally, participants responded to the same manipulation check questions as in the previous studies as well as two additional questions assessing the effectiveness of the manipulation of the variance in the ratings of the other-subcategory brand: specifically, participants responded to “Which of the two hotels had a greater share of ‘terrible’ ratings?” and “Which of the two hotels had less consistent ratings?”. Both questions were anchored at 1 = “Definitely the DOWNTOWN HOTEL (A member of the LL Hotel Group), 4 = “Equal share of ‘terrible’ ratings”/ “Equally consistent ratings”, and 7 = “Definitely the QUALITY HOTEL (An Independent Hotel)”.      51 10.2 Results  Manipulation check. A two-way ANOVA on the mean of the fourteen THS manipulation check items (α = .99) with feelings toward the focal brand and review variance as between-subjects factors only produced a main effect for feelings such that participants in the hatred conditions experienced more hatred toward the Waterside Inn than those in the indifference conditions (Mhatred = 6.56, SD  = 1.42 vs. Mindifference = 1.84, SD  = 1.42, F(1, 391) = 986.28, p < .001). Neither the effect of review variance (F < 1) nor the interaction (F(1, 391) = 3.05, p = .08) were significant. With respect to the manipulation of the variance in the reviews of the other-subcategory brand, I found only a significant effect of review variance such that participants in the higher variance conditions perceived the other-subcategory brand to have a greater share of bad ratings (Mhigher variance = 2.38, SD = 1.99 vs. Mequal variance= 4.09, SD  = .65, F(1, 391) = 128.00, p < .001) and less consistent ratings (Mhigher variance= 2.77, SD = 1.85 vs. Mequal variance = 4.05, SD = .61, F(1, 391) = 81.63, p < .001). No other effects were significant (all Fs < 1). Purchase likelihood. I used a mixed model ANOVA to assess how feelings (hatred vs. indifference) and the variance in the ratings of the other-subcategory brand (low vs. high) affect the purchase likelihood for the two competitors. This analysis produced a significant three-way interaction between feelings, review variance, and purchase likelihood (F(1, 391) = 5.41, p < .001, η2p = .014, see figure 4). The two-way interactions between feelings and purchase likelihood (F(1, 391) = 11.74, p < .001, η2p = .029) and between review variance and purchase likelihood were also significant (F(1, 391) = 54.61, p < .001, η2p = .123). The main effect of purchase likelihood was not significant (F(1, 391) = 1.74, p > .18).     52 Figure 4. Purchase Likelihood for Chain and Independent Hotels (Study 5) Panel A: Equal Variance in the Ratings for the Other-Subcategory Brand  Panel B: Higher Variance in the Ratings for the Other-Subcategory Brand	   Note: The error bars represent 95% confidence intervals   In line with predictions, brand hatred induced preference shifts leading consumers to prefer the brand from another subcategory to one from the same subcategory when the variance in the ratings of the other- and same-subcategory brand was the same (i.e., replication condition). In the hatred condition, consumers exhibited a significant preference for the brand from the other subcategory (Mother-subcategory = 6.58, SD = 2.13 vs. Msame-subcategory = 4.51, SD = 2.33, F(1, 391) = 37.69, p < .001, η2p = .088; see figure 4, panel A). When participants were indifferent, they did   53 not prefer one brand to the other (Mother-subcategory = 6.07, SD = 1.69 vs. Msame-subcategory = 6.01, SD = 1.71, F < 1; see figure 4, panel A).   The patterns differed when the ratings of the other-subcategory brand were highly variable and this brand thus was not well suited to assuage concerns about self-protection. In this case, consumers exhibited a significant preference for the same-subcategory brand in the hatred condition (Mother-subcategory = 5.02, SD = 2.26 vs. Msame-subcategory = 6.35, SD = 2.13, F(1, 391) = 13.85, p < .001, η2p = .034; see figure 4, panel B). Participants in the indifference condition also preferred the same-subcategory brand (Mother-subcategory = 4.88, SD = 2.36 vs. Msame-subcategory = 6.59, SD = 2.14, F(1, 391) = 25.71, p < .001, η2p = .062; see figure 4, panel B).  10.3 Discussion Using a moderation-of-process approach, study 5 corroborates that self-protection concerns underlie the preference shift induced by brand hatred. In the absence of competing cues for evaluating the safety of the options in the consideration set, hatred again led to a preference for other-subcategory competitors. However, in the presence of a cue that directly conflicts with self-protection concerns – more variability in the ratings for the other-subcategory brand – hatred produced a preference for the same-subcategory brand. Notably, by demonstrating that consumers readily use variance in ratings to assess the safety of different options, I identify an instance in which variance in ratings reliably leads to decreased preferences for an option, extending previous work on consumer responses to variance in ratings.    54 Chapter 11: Study 6 Study 6 provided further evidence for my self-protection account by examining the efficacy of a strategy managers might use to shield their brands from negative effects resulting from hatred toward close competitors. A natural response for managers might be to believe that having a superior quality reputation suffices to protect their brands from negative effects resulting from hatred for same-subcategory competitors. However, in line with my proposed mechanism, I predict that a same-subcategory brand can better insulate itself when its superior quality perceptions are paired with marketing instruments that assuage consumers’ self-protection concerns. In this study, I compared the effectiveness of quality signals (i.e., a higher rating) alone versus quality signals combined with money-back guarantees (a means that can directly address consumers’ self-protection concerns; Hogreve and Gremler 2009) in mitigating the negative influence of brand hatred for a same-subcategory competitor. The study used a 2 (feeling: hatred vs. indifference; between-subjects) x 3 (competitor advantage: none vs. quality vs. quality plus guarantee; between-subjects) x 2 (purchase likelihood: same subcategory competitor vs. different subcategory competitor; within-subject) mixed-design.   11.1 Method Participants and design. Three hundred and twenty-nine participants recruited from Amazon Mechanical Turk (59% female, Mage = 33.9) read the hotel scenario. In study 6, all participants imagined staying at the chain hotel. Like in the previous studies, I manipulated feelings toward this brand. In this study, I additionally altered the information provided about the two brands on the booking portal screenshot in order to test the efficacy of quality signals and money-back guarantees (see appendix A16). Participants were randomly assigned to one of three   55 levels of the competitor advantage factor: (1) no advantage, in which the same subcategory competitor had the same quality rating as the other-subcategory brand; (2) quality advantage, in which the same-subcategory brand had a higher rating ("8.0", "Very Good") than the brand from the other subcategory; or (3) quality advantage plus guarantee, in which the higher quality same-subcategory brand (rating of 8.0) offered a 100% satisfaction money-back guarantee. I used the same dependent variable and manipulation check questions as in studies 3 and 4.   11.2 Results Manipulation check. A two-way ANOVA on the mean of the hatred items (α = .89) with feelings toward the focal brand (hatred vs. indifference) and same-subcategory competitor advantage (no advantage vs. quality advantage vs. quality advantage plus money-back guarantee) as between-subjects factors only produced a main effect for feelings toward the focal brand such that participants in the hatred conditions experienced more hatred toward the Waterside Inn (Mhatred = 36.21, SD  = 24.87 vs. Mindifference = 7.79, SD  = 13.69, F(1, 325) = 135.72, p < .001). No other effects were significant (all Fs < 1, ps > .34).  Purchase likelihood. I used a mixed model ANOVA with feelings toward the focal brand and same-subcategory competitor advantage as independent variables and purchase likelihood for same- and other subcategory brand as the repeated dependent measure. Both two-way interactions were significant (Purchase Likelihood x Same-Subcategory Competitor Advantage: F(2, 323) = 15.45, p < .001, η2p = .087; Purchase Likelihood x Feelings toward the focal brand: F(1, 323) = 15.63, p < .001, η2p = .046, see figure 5). These interactions qualified the main effect of brand (F(1, 323) = 15.01, p < .001, η2p = .044). The three-way interaction was not significant (F(2, 323) = 1.10, p > .33).    56 Figure 5. The Effect of Different Intervention Strategies on Consumer Preferences (Study 6) Panel A: Feeling toward Focal Brand = Hatred  Panel B: Feeling toward Focal Brand = Indifference	   Note: The error bars represent 95% confidence intervals.  Contrast analyses in the hatred condition suggested that when the same-subcategory competitor did not have a quality advantage, consumers exhibited a significant preference for the brand from the other subcategory (Mother = 6.98, SDother = 1.52 vs. Msame = 5.75, SDsame = 2.24, F(1, 323) = 11.91, p < .001, η2p = .036; see figure 5, panel A), replicating the pattern documented in the previous studies. Results further revealed that if the same-subcategory competitor had a quality advantage, the preference for the other-subcategory brand was eliminated but not   57 reversed (Mother = 6.75, SDother = 1.89 vs. Msame = 6.61, SDsame = 1.82, F < 1, p = .70; see figure 5, panel A). However, when the brand from the same subcategory with a quality advantage also offered a money-back guarantee it was not negatively affected by hatred and was significantly preferred by consumers (Msame = 7.67, SDsame = 1.90 vs. Mother = 6.35, SDother = 1.92, F(1, 323) = 11.21, p < .001, η2p = .034; see figure 5, panel A). In contrast, when participants were indifferent to the Waterside Inn they preferred the higher-quality same-subcategory option regardless of whether it offered a money-back guarantee (Msame = 7.71, SDsame = 1.43 vs. Mother = 5.79, SDother = 1.82, F(1, 323) = 33.01, p < .001, η2p = .093) or not (Msame = 7.13, SDsame = 1.97 vs. Mother = 5.98 SDother = 1.80, F(1, 323) = 9.86, p < .01, η2p = .026; see figure 5, panel B).  11.3 Discussion Study 6 examined how brands might effectively react to hatred for same-subcategory competitors. Having a reputation for superior quality by itself is insufficient to fully insulate a brand from the negative effects of hatred for its close competitors. Only when high-quality brands utilize tools like money-back guarantees, which directly assuage consumer concerns about self-protection, they remain fully untarnished when consumers hate close competitors.  To further bolster the self-protection based account for the effect of hatred on consumer preferences I also included a measure of self-protection concerns that asked participants to rate the relative safety of the options in the choice set rather than consumers’ general concerns about protecting the self (as in the measure in studies 4a and 4b). Specifically, participants indicated which of these two hotels was the safer option to pick in order to minimize the possibility of experiencing unpleasant or negative events during their stay (with the two hotels serving as the endpoint on this scale). A moderated mediation analysis described in detail in appendix A17 with   58 this item as the mediator, produced patterns consistent with the central prediction that only the combination of superior quality with a money-back guarantee was sufficient in assuaging consumers’ concerns about self-protection.      59 Chapter 12: Study 7 In my final study, I examined a potential boundary condition for the preference shifts induced by hatred. Prior work has suggested that hatred might produce aggressive motivations against the focal hated object (Rempel and Burris 2005), which then motivate consumers to engage in behaviors intended to harm a brand (Kähr et al. 2016). For instance, Nasr Bechwati and Morrin (2003) found across two studies that consumers experiencing a high desire for vengeance—a feeling related to hatred—were more likely to switch to a suboptimal, but more direct, same-subcategory rival (12% vs. 0% when vengeance was low).  In previous studies presented in this dissertation, I mirrored common choice contexts (e.g., when searching for hotels on TripAdvisor) wherein consumers have limited information about the relationship between the focal brand and the other brands in the choice set. The results of study 4b suggest that merely making the choice observable to the focal hated brand does not suffice to attenuate the preference for a brand from a different subcategory. However, there can be cues in the marketplace that highlight the relationship between brands, such as in comparative advertising or media stories about rivalries. As outlined in my third hypothesis, I predict that if the rivalry between same-subcategory competitors is made salient the effect of hatred on preference for the other-subcategory brand will be attenuated. In this case, the motivation to take revenge against the focal brand by choosing a more direct same-subcategory rival and the need for self-protection conflict; hence, I no longer expect a preference for the other-subcategory brand as a consequence of hatred. To increase the comparability between my work and prior research (e.g., Nasr Bechwati and Morrin 2003), I test these predictions in an industry that frequently triggers hatred (see qualitative research) and was used by existing research: mobile phone operators.    60 Across studies I have shown that brand hatred leads consumers to eschew brands from the same subcategory, unless there are diagnostic cues available that cast doubt on the safety of the other-subcategory brand (e.g., a greater variance in consumer ratings, study 5). Prior research suggests that the degree to which a brand is established (vs. new) might be another cue to assess the safety of an option (Griskevicius and Kenrick 2013). In this study, I pit these cues (i.e., the degree to which it is established vs. category membership) against each other such that the brand from the other subcategory is the newer option whereas the brand from the same subcategory is more established. If hatred still produces a preference for the brand from the other subcategory, it provides a conservative test of the primacy of category membership. Finally, in this study I return to a binary choice variable, which more closely mimics the decision involved in switching mobile phone operators. Participants were assigned to one of two conditions, which varied the salience of the competitive rivalry (i.e., low vs. high). Different from my previous studies, all participants read about a hatred-evoking experience with the firm.  12.1 Method Participants and design. Three hundred and three participants (58% female, Mage = 38.4), recruited from Amazon Mechanical Turk, imagined being in Canada for an extended period of time for work and read a scenario about their experiences with a local mobile operator named Fido, an established brand. I modeled my paradigm after Nasr Bechwati and Morrin (2003) and presented participants with a newspaper article about trends in the country’s telecommunications industry. The article described the industry structure, which was comprised of two categories, established firms and newer companies. In addition, the article contained information on the competitive rivalry between the focal brand, Fido, and Telus, the other established company in   61 the industry (see appendix A18 for the full stimuli). In the high salience condition, the article explicitly stated: “While there are high levels of competition in this industry, the two key rivals in the industry are the two established operators, Fido and Telus, which fiercely battle for customers. This rivalry has become increasingly heated as the CEO of Fido has repeatedly emphasized that the key strategic priority of Fido is taking customers from Telus.” In contrast, in the low salience condition, the article concluded as follows: “While there are high levels of competition in this industry, the most direct competitors in the industry are Fido and Telus, which are both established companies.”  Next, participants were told that reading the article reminded them of their hatred-evoking experiences with Fido: they had faced repeated unfair charges and mistreatment from customer service employees. I modeled the scenario on the descriptions of one of my informants from the in-depth interviews. Regardless of whether the competitive rivalry was salient or not, I told participants that they had decided to switch brands. Subsequently, I measured the dependent variable by telling participants that they need to inform their current provider, Fido, to which competitor they were switching in order to carry over their current number to the new operator (0 = Wind [the newer, i.e., other-subcategory brand], 1 = Telus).  Finally, I assessed the effectiveness of my manipulation of competitive salience by asking participants to respond to the question “which two companies battle MOST fiercely for customers?” (1 = “Fido and Wind”; 2 = “Fido and Telus”; 3 = “Wind and Telus”; 4 = “All companies battle each other equally”). Participants also indicated which two companies they perceived to be most similar in the local telecommunications industry (1 = “Fido and Wind”; 2 = “Fido and Telus”; 3= “Wind and Telus”). I predict that in the high (vs. low) salience condition a greater share of consumers perceives the same-subcategory brands as more fierce competitors,   62 however perceptions of similarity should be unaffected by the competitive salience manipulation. Participants also responded to the same THS manipulation check items as in the previous studies and provided demographic information.  12.2 Results  Manipulation check. To assess the effectiveness of my manipulation of the salience of the rivalry I ran a logistic regression on the perceptions of the same-subcategory brands as the closest rivals (0 = no, 1 = yes) with salience of the rivalry as the independent variable. This analysis produced a significant effect for the salience of the competitive rivalry (βexp = 1.61, SE = .31, Wald χ2(1) = 27.56, p < .001). When salience of the rivalry was high, participants were significantly more likely to perceive Fido and Telus as the fiercest rivals (88.7%) than when it was low (61.2%). However, the salience of the brand rivalry did not affect similarity perceptions: across conditions participants perceived Fido and Telus to be the most similar firms (low: 98.7% vs. high: 98.7%; χ2 (1) < 1). Levels of hatred on the THS scale (α = .94) did not differ between the two salience conditions (Mlow = 7.05, SD  = 1.42 vs. Mhigh  = 6.91, SD  = 1.39, t < 1). Choice of provider. Using a binary logistic regression to test how the salience of the competitive rivalry (0 = control, 1 = high) affected preferences for the new provider (0 = Wind; other subcategory; 1 = Telus; same subcategory), I found a significant positive effect of salience on the choice of the Telus, the same-subcategory brand (βexp = .90, SE = .24, Wald χ2(1) = 14.59, p < .001). In the low salience condition, I found a significant preference for Wind, the brand from the other-subcategory, replicating the previous studies (Choice Shareother-subcategory = 63.8%). This preference was significantly greater than chance (χ2(1) = 11.61, p < .001, d = .58). In contrast, when the salience of the rivalry was high, participants exhibited a preference for Wind,   63 the closer rival of Fido from the same subcategory (Choice Sharesame-subcategory = 58.3%), which was significantly greater than chance (χ2(1) = 4.14, p = .042, d = .34) as well.  12.3 Discussion The final study of my dissertation illuminates how the salience of the competitive rivalry between the focal and a close, same-subcategory competitor moderates the effect of brand hatred on preferences for competing brands. While I replicate the effect of hatred on preferences in the control condition, I found a preference reversal when the salience of the brand rivalry is high. This finding also highlights hatred’s emotionally pluripotent nature, as choosing a fierce competitor might at least in part be driven by the anticipated joy derived from hurting the hated brand. This finding reinforces the potency of highlighting competitive salience for influencing consumers’ brand preferences (Paharia et al. 2014), particularly when consumers experience hatred.       64 Chapter 13: General Discussion In this dissertation, I tested how hatred for a brand affects consumer preferences toward competitors. Converging evidence from qualitative, correlational, experimental, and field data suggests that hatred for a brand negatively affects same-subcategory competitors. The qualitative study highlighted how the experience of brand hatred motivates consumers to avoid further brand harm and the steps they take in order to do so. Study 1 corroborated the qualitative study by using online reviews of brand experiences. Results revealed that hatred-evoking experiences reduce the propensity of consumers to review other members of the same subcategory and increase the duration before they visit and review same-subcategory providers. Study 2 replicated these patterns using a correlational approach in multiple categories. Study 3 documented that hatred leads consumers to choose providers from a different subcategory and demonstrated that these preference shifts do not emerge when consumers are indifferent or dissatisfied. Studies 4a, 4b, and 5 identified, through mediation and moderation, the underlying role of consumer concerns for self-protection in driving the preference shift. Study 6 provided further evidence for my self-protection account by examining strategies for brands whose close rivals are hated. The final study showed that making the competitive rivalry salient, offering an opportunity for a consumer to inflict harm to a hated brand, produces a preference for closer, same-subcategory rivals.   13.1 Theoretical Contributions From a theoretical perspective, my dissertation contributes to the literature on how consumers relate to brands. Hitherto, research on consumer-brand bonds has largely focused on strong, positive engagement with brands (e.g., Fournier 1998, 2009; Thomson, MacInnis, and   65 Park 2005; Yim, Tse, and Chan 2008), documenting how such engagement creates positive outcomes for the focal brand such as engagement in positive word-of-mouth, brand loyalty, and greater share-of-wallet (e.g., Batra, Ahuvia, and Bagozzi 2012; Park et al. 2010; Sprott, Czellar, and Spangenberg 2009). In many marketplaces, however, the majority of consumers may not form strong, positive bonds with brands (e.g., Alvarez and Fournier 2016; Mende, Bolton, and Bitner 2013). Instead, they may develop strong, extremely negative feelings toward brands, such as the focus of this dissertation: hatred. Even beyond the consumer realm, hatred remains critically understudied (Harris 2018). The sparse empirical investigations that do exist exclusively focus on the effects of hatred on the focal object (e.g., Kähr et al. 2016) or social group (e.g., Halperin 2008). My dissertation extends this line of research by exploring hatred's potency to create effects beyond the focal actor (i.e., brand) that generated this feeling. Specifically, hatred influences perceptions and choices of competing brands in choice sets that do not even feature the focal brand. In other words, consumer relationships with brands do not operate in a vacuum. Given its focus on how consumers become deeply connected and engaged with brands (e.g., Batra et al. 2012; Yim et al. 2008), past research has, at least implicitly, implied that the connection a consumer builds with a focal brand has uniform effects on competing brands. For instance, if consumers come to love a focal brand, they are likely to entirely disregard all other competing brands paying little attention to the particular relationships and constellations between the focal brand and its competitors. My dissertation highlights the need to extend frameworks of how consumers form relationships with brands to incorporate more negative consumer-brand connections and account for their effects beyond a focal brand. Unlike consumers who positively bond with a brand, consumers who experience hatred for a brand are likely in the market for alternatives. My results highlight the importance of various   66 aspects of the marketplace relationships between different brands (e.g., category membership, intensity of rivalry) in providing meaning for consumers and shaping their preferences when they hate a focal brand. I also demonstrate that brand hatred, but not a less extreme feeling like dissatisfaction, produces a robust preference for an other-subcategory brand when considering alternatives to the focal brand. This finding is notable since existing research documents that many negative feelings produce similar intentions and behaviors toward the focal brand (Romani et al. 2012). Such responses toward focal brands include complaining, negative word-of-mouth, and the desire to switch to a different firm. Thus, while different negative feelings such as hatred and dissatisfaction might prompt consumers to switch to rival brands, only hatred consistently led to a preference for a particular type of competitor, namely a less close competitor from a different subcategory. More generally, my dissertation not only demonstrates that hatred differs from other sentiments toward brands, but also identifies the consumption arena as a promising domain to further explore hatred’s nature and consequences.  My work is the first to highlight that in certain instances, such as when choosing brands, hatred not only triggers vengeful sentiments (van Doorn 2018), but also concerns about protecting the self (see also Roseman and Steele 2018). Past consumer research has documented how hatred can produce revenge motivations and that these motivations in turn can lead consumers to avoid (Grégoire et al. 2009) and sabotage (Kähr et al. 2016) focal hated brands. Replicating previous research I also find in study 4b that hatred indeed leads to increased revenge motivations. However, while revenge appears to be crucial for shaping consumer behavior toward a focal brand (e.g., negative word-of-mouth, vindictive complaining, marketplace aggression), my results suggest that self-protection concerns are generally more   67 influential in the formation of preferences in subsequent consumption situations that do not directly involve the focal brand. Despite its prevalence and potency, empirical investigations of hatred remain scarce and the empirical literature to date has focused primarily on the aggressive action tendencies toward the focal object (e.g., Halperin 2008; Harris 2018; Kähr et al. 2016). A central emotivational goal at the heart of hatred is the removal or elimination of the hated other from one’s environment (Fischer et al. 2018; Roseman and Steele 2018; Sternberg and Sternberg 2008). My findings suggest that the path to achieving this goal – at least in the consumption domain – typically involves eschewing entities (i.e., brands) that resemble the hated object rather than deliberately acting in ways that would further the destruction of the hated object (e.g., choosing a closer, more direct competitor).  By identifying hatred as triggering concerns about protecting the self, I also add to an emerging stream of research documenting conditions under which such concerns become activated (Lisjak and Lee 2014; Winterich et al. 2014). In the context of brand choices, hatred-evoking consumption experiences increase generalized concerns about protecting the self and lead consumers to prefer brands that are perceived as safer. My results suggest that concerns about protecting the self manifest in consumer choice via a preference for brands from a different subcategory rather than through other avenues suggested in prior research such as a more careful examination of available options or a preference for more established brands (Griskevicius and Kenrick 2013).  Finally, at the intersection between theory and methodology, my first study offers a novel way of utilizing data from review platforms. Typically, prior research has focused on how certain elements of the consumption experience influence the star ratings of consumers (e.g., Paharia et al. 2014; Yang and Aggarwal 2018) and the language consumers use in their reviews (e.g., Chen   68 2017; Henkel et al. 2018). A second complementary line of research has focused on the elements in reviews that render them useful and helpful in the eyes of other users (e.g., Chen and Lurie 2013; Elder et al. 2017). Going beyond these approaches, my dissertation exploits the substantial level of engagement of many reviewers on Yelp over time to examine how certain experiences (i.e., potentially hatred- vs. indifferent restaurant experiences) influence the subsequent (reviewing) behavior of users. Focusing on such within-user reviewing patterns and temporal variation appears to be a promising avenue for testing a variety of consumer research theories.  13.2 Managerial Implications The findings of this dissertation emphasize the importance for brand managers to be cognizant and attentive toward consumers' feelings for their competitors. As evidenced by Spirit Airlines and the survey of managers discussed in the introduction, conventional managerial wisdom holds that hatred toward a close competitor is beneficial or at least not harmful for their own brand. The empirical results suggest that this may be wishful thinking, as I consistently find across different categories and methods that brand hatred tends to negatively affect close competitors of the hated brand. It appears pivotal for managers to monitor, anticipate, and potentially react to consumer feelings toward competing brands in order to maximize outcomes for their own brand. Managers should keep track of changes in consumer feelings by monitoring social media sites and online review sites for spikes in reports of extremely negative feelings such as hatred resulting from consumer experiences with competing brands. If these brands are close, same-subcategory competitors, brand managers should consider supplementing or modifying their offerings such that consumers do not perceive their brand as less safe than brands from more distant subcategories. For instance, on their online channels, it might be   69 relatively easy to adapt offerings (e.g., the descriptions on booking portals) to include some form of guarantee or other self-protection related claim if a same-subcategory competitor is attracting significant levels of hatred. Alternatively, managers might build on the results of study 6 and allow consumers to customize their purchases by offering specific add-ons (e.g., insurances, guarantees) that can make consumer feel safer. With respect to offline marketing activities, managers should consider employing marketing tactics that are high in credibility (e.g., reporting the exact position on a ranked list, provide detailed ratings of the brand on online review platforms) in order to bolster the trustworthiness and thereby the perceived safety of their brand (Isaac and Grayson 2017). The results of study 7 suggest yet another, more provocative strategy to deal with hatred for close competitors: communicating and emphasizing the fierce rivalry with the close, same-subcategory competitor that is the target of consumers’ hatred. There are a few examples of corporate leaders that seem to endorse this strategy. Maybe the most colorful example of this approach is John Legere, the chief executive officer of T-Mobile US. He regularly taunts close competitors (some of whom arguably hated) on social media and even went so far as presenting his approach to “trash-talking rivals” in a Harvard Business Review feature article (Legere 2017).  My results also suggest that brand hatred is generally beneficial for brands that do not belong to the same subcategory as the hated brand. Across studies I find that these effects emerge in the absence of any actions by these competing brands. While speculative, it is possible that such effects might be further magnified by highlighting how the other-subcategory brand's offerings differ from those of the hated brand. Thus, the results suggest that when deciding whether and how to reference competitors and category membership in a brand's advertising, managers should not only consider factors related to the competitive relationships such as   70 relative size (Paharia et al. 2014) or relative quality (Kim and Tsai 2012), but perhaps also whether there are strong, negative feelings such as hatred toward competing brands.   13.3 Future Research and Concluding Remarks As with any research this work has limitations, which present avenues for future research. While hatred is often viewed as illegitimate by scholars and laypersons alike (Ben-Ze'ev 2001; Halperin 2012), it does not only appear to be prevalent in marketplaces, but individuals seem more willing to acknowledge that they experiences hatred for an impersonal entity such as a brand (vs. for another person or social group). In my dissertation, I focused on exploring how hatred for a brand influences subsequent preferences for competing brands, one of, if not the most, important outcomes of brand hatred from the perspective of managers and firms. There is, however, a need for more work to better delineate the affective space of hatred (Rempel et al. 2018; Roseman and Steele 2018). As an initial effort to this end, I conducted a follow-up study (for a presentation of the full stimuli see appendix A19) in which I directly compared how anger (without hatred) and hatred affected preferences for competing brands. Results suggest that hatred, but not anger led a significantly greater propensity to choose a more distant competitor in order to protect the self from further harm rather than choosing a close competitor in order to get back at the focal brand. While the emotional components of hatred – contempt, disgust, and anger – often empirically cluster together (for a recent review see Gervais and Fessler 2017), it appears that the consumption domain might be a fertile ground to enhance our understanding of the structural configuration of hatred in general. Potentially, future research might also use the hatred consumers feel for brands via other methodologies such as asking participants to project their feelings of hatred onto other social objects (e.g., Cohen, Pham, and Andrade 2008; Sengupta, Dahl, and Gorn 2002). I explored the feasibility of such a projection technique   71 approach (Fisher 1993) in a conceptual replication of study 3 wherein participants were asked to project their actual feelings (i.e., hatred, dissatisfaction, or indifference) for a brand onto a fictional brand in a more controlled choice setting (i.e., the hotel paradigm used in studies 4-6). Such methodological approaches are well suited to use the hatred consumers experience for brands to further explore our understanding of hatred as a psychological construct. A worthwhile avenue for future research would be to explore when hatred produces similar versus divergent action tendencies and behaviors when compared to the singular experience of its components (i.e., anger, contempt, and disgust) in other important domains beyond brand preferences (e.g., interpersonal, romantic, or work relationships). My dissertation focuses primarily on consumers’ own real or imagined experiences as the cause of their feelings toward brands. However, my self-protection account might also be relevant for situations in which feelings arise as a result of circumstances other than personal experience with a brand such as in scandals. It would be interesting to examine when brand scandals are strong enough to evoke hatred and in turn activate consumers’ need for self-protection (e.g., Borah and Tellis 2016; Roehm and Tybout 2006). Most of the brand scandals and harm crisis literature has focused on more cognitive processes rooted in the accessibility–diagnosticity framework (Feldman and Lynch 1988) emphasizing that same-subcategory brands are more likely to be negatively affected when the transgressing company is typical of its category (e.g., Burger King for hamburgers) and when the scandal attribute is closely associated with the focal category (e.g., a problem with its hamburger meat) rather than loosely (e.g., a problem with its ice cream). In my studies, the transgressions that give rise to hatred are either similar (study 1) or identical (study 4a) across the different categories. Brand typicality did not emerge as a prerequisite for hatred’s effect on consumers’ preference for other brands, but it   72 would be interesting to better understand how brand characteristics relate to hatred and its consequences. To provide support for the self-protection account, I employ mediation (studies 4a and 4b) and moderation-of-process (studies 5–7) approaches. Yet, one alternative and potentially more parsimonious account for the effects of brand hatred on consumers’ preferences for competing brands is consumer learning. Viewed from this perspective, exposure to and experiences with the focal brand lead consumers to encode information not only about this brand, but also about (close) competitors and categories (e.g., Erdem and Keane 1996; Hoch and Deighton 1989). This encoded information in turn shapes consumer preferences. While it is important to acknowledge the role of consumer learning in the formation of consumer preferences after a hatred-provoking consumption experience, a learning account cannot fully account for the findings presented in my dissertation. First, it does not predict the preference reversals demonstrated in study 5 when an other-category brand is characterized by a cue (i.e., higher variability in consumer ratings) that reduces its capability to address consumers’ concerns about self-protection. Second, a learning account is agnostic to hatred’s motivational properties and therefore does not predict that the effect of brand hatred is moderated by the salience of the competitive rivalry, which reduces the salience of concerns about self-protection, as demonstrated in study 7. Finally, a learning account predicts moderation by familiarity such that the impact of a hatred-evoking consumption experience should be most pronounced for those with limited prior category experience (Hoch and Deighton 1989). However, I did not find evidence for moderation by familiarity in my analyses of the Yelp data reported in study 1. Taken together, consumer learning contributes to preference formation after hatred evoking   73 brand experiences, however concerns about self-protection are the more proximal force in shaping preference for competing brands after hatred evoking brand experiences. Across studies, the experiences that generated hatred for a brand ranged from performance- (e.g., rude employee behavior, poor product performance) to morality-related violations (e.g., corporate greed, unhygienic servicescapes). However, consumers may develop hatred for brands for reasons unrelated to actions by the brand, such as symbolic identity incongruence. Identity-related hatred may emerge when brands are used by negative reference groups held in contempt by consumers (e.g., Berger and Heath 2008). This type of hatred may lead to concerns other than self-protection, resulting in different outcomes for competing brands. My findings suggest that consumers rely on category membership to choose between different options. In order to increase experimental control, I provided clear category markers. However, there are plenty of consumption situations in which multiple aspects could be used to categorize brands. For instance, a consumer choosing between different European airlines might organize brands based on country-of-origin (e.g., British) or class of service (e.g., low-cost vs. full-service). Depending on the attribute used to categorize brands, a brand may either belong to the same or a different subcategory. It would be interesting to examine whether hatred (vs. other negative feelings) changes how people categorize brands and other objects.  A related question worthy of investigation is how brand hatred affects subsequent consumption when consumers conceive all brands as belonging to the same subcategory. Failing to perceive sufficiently differentiated categories could either be a function of the absence of actual differentiation between brands or a more subjective perception on the behalf of the consumer. In either situation, it would be difficult for consumers to assuage concerns about self-protection in the fashion outlined in this dissertation. When the consumption of the product or   74 service offered by the hated brand is discretionary, hatred might spur choice deferral and category exit (Mochon 2013). However, it less clear how consumers might seek to protect themselves when the purchases are non-discretionary. Past research has suggested that one way consumers might deal with such potentially distressing decisions is to postpone consumption (e.g., Chen and Pham 2019). The results of my study 6 are suggestive of another potential strategy consumers might employ, namely purchasing add-ons (e.g., insurances, service guarantees) that can provide a layer of protection. A third way how hateful consumers might deal with such aversive choices might be to delegate the this difficult, dreaded decision to others (Steffel and Williams 2018). Thus, elucidating the aspects of choice situations that determine which of these strategies consumers rely on when brands are lumped together into one category is an exciting avenue for future research. Finally, across studies I focused on how hatred for a brand affects preferences for competing brands in the same product or service category as the focal hated brand. Supplementary analyses decomposing the preference shifts induced by brand hatred suggested that the effects of brand hatred were primarily driven by a reduction in the preference for the same-subcategory option in the choice set. Preferences for the brand from the other subcategory remained relatively constant. This implies that consumers eschew brands from the same subcategory or brands with the other cues suggesting greater potential harm (e.g., greater variability in ratings, study 5). These results further substantiate to the central role of self-protection in shaping consumption within the same product or service category as the focal hated brand. It is important to note that I found in studies 4a and 4b that hatred led to increased generalized concerns about protecting the self. This raises questions about the behavioral outcomes of brand hatred in other domains. My qualitative research produced a striking example   75 wherein one of the respondents highlighted how his hatred for a brand made him more hesitant and cautious in other unrelated domains of life (i.e., deciding about whether to propose to his girlfriend). Such spillovers between domains dovetail with prior research linking self-protective concerns to loss aversion (Li et al. 2012). One might speculate whether individual differences in the chronic activation of self-protection motives (Neel et al. 2016) influence the reach of (brand) hatred beyond the focal (consumption) category. More work is needed to unravel the conditions under which hatred for a brand or other objects affects decisions in unrelated domains.   To conclude, my dissertation highlights the importance of brand hatred and its far-reaching effects beyond a focal brand. Managerial wisdom as exemplified in my opening example from Spirit suggests that hatred for competitors is beneficial for a brand. The results of my dissertation caution against such generalized enthusiasm as brand hatred increases concerns about self-protection, which in turn reduce rather than increase consumers’ propensity to choose close, same-subcategory competitors of a hated brand. More generally, despite its potency and prevalence, empirical examinations of hatred remain scarce. I hope that my dissertation further invigorates an emerging recognition of hatred as psychological construct worthy of study. As discussed in my future research section, consumer-brand connections are not only a promising domain for studying hatred, but present numerous opportunities for unpacking the complexity of experiencing hatred that can deepen our understanding of its nature and potency.      76 References  Alvarez, Claudio and Susan Fournier (2016), "Consumers’ Relationships with Brands," Current Opinion in Psychology, 10 (August), 129-35. Arnould, Eric J. and Melanie Wallendorf (1994), "Market-Oriented Ethnography: Interpretation Building and Marketing Strategy Formulation," Journal of Marketing Research, 31 (4), 484-504. Arsel, Zeynep (2017), "Asking Questions with Reflexive Focus: A Tutorial on Designing and Conducting Interviews," Journal of Consumer Research, 44 (4), 939-48. Batra, Rajeev, Aaron Ahuvia, and Richard P. Bagozzi (2012), "Brand Love," Journal of Marketing, 76 (2), 1-16. Baumeister, Roy F. and David A. Butz (2005), "Roots of Hate, Violence, and Evil," in The Psychology of Hate, ed. Robert J. Sternberg, Washington, DC: American Psychological Association, 87-102. Ben-Ze'ev, Aaron (2001), The Subtlety of Emotions, Cambridge, MA: The MIT Press. ______ (2018), "Is Hate Worst When It Is Fresh? The Development of Hate over Time," Emotion Review, forthcoming.   77 Berger, Jonah and Chip Heath (2008), "Who Drives Divergence? Identity Signaling, Outgroup Dissimilarity, and the Abandonment of Cultural Tastes," Journal of Personality and Social Psychology, 95 (3), 593-607. Berger, Jonah and Katherine L. Milkman (2012), "What Makes Online Content Viral?," Journal of Marketing Research, 49 (2), 192-205. Borah, Abhishek and Gerard J. Tellis (2016), "Halo (Spillover) Effects in Social Media: Do Product Recalls of One Brand Hurt or Help Rival Brands?," Journal of Marketing Research, 53 (2), 143-60. Bougie, Roger, Rik Pieters, and Marcel Zeelenberg (2003), "Angry Customers Don't Come Back, They Get Back: The Experience and Behavioral Implications of Anger and Dissatisfaction in Services," Journal of the Academy of Marketing Science, 31 (4), 377-93. Carver, Charles S. and Eddie Harmon-Jones (2009), "Anger Is an Approach-Related Affect: Evidence and Implications," Psychological Bulletin, 135 (2), 183-204. Chen, Charlene Y. and Michel Tuan Pham (2019), "Affect Regulation and Consumer Behavior," Consumer Psychology Review, 2 (1), 114-44. Chen, Zoey (2017), "Social Acceptance and Word of Mouth: How the Motive to Belong Leads to Divergent Wom with Strangers and Friends," Journal of Consumer Research, 44 (3), 613-32.   78 Chen, Zoey and Nicholas H. Lurie (2013), "Temporal Contiguity and Negativity Bias in the Impact of Online Word of Mouth," Journal of Marketing Research, 50 (4), 463-76. Cohen, Joel B., Michel Tuan Pham, and Eduardo B. Andrade (2008), "The Nature and Role of Affect in Consumer Behavior," in Handbook of Consumer Psychology, ed. Curtis P. Haugtvedt, Paul M. Herr and Frank R. Kardes, New York; London: Lawrence Erlbaum Associates, 297-348. Crossley, Craig D. (2009), "Emotional and Behavioral Reactions to Social Undermining: A Closer Look at Perceived Offender Motives," Organizational Behavior and Human Decision Processes, 108 (1), 14-24. Darke, Peter, Laurence Ashworth, and Kelley Main (2010), "Great Expectations and Broken Promises: Misleading Claims, Product Failure, Expectancy Disconfirmation and Consumer Distrust," Journal of the Academy of Marketing Science, 38 (3), 347-62. Elder, Ryan S., Ann E. Schlosser, Morgan Poor, and Lidan Xu (2017), "So Close I Can Almost Sense It: The Interplay between Sensory Imagery and Psychological Distance," Journal of Consumer Research, 44 (4), 877-94. Erdem, Tülin and Michael P. Keane (1996), "Decision-Making under Uncertainty: Capturing Dynamic Brand Choice Processes in Turbulent Consumer Goods Markets," Marketing Science, 15 (1), 1-20.   79 Feldman, Jack M. and John G. Lynch (1988), "Self-Generated Validity and Other Effects of Measurement on Belief, Attitude, Intention, and Behavior," Journal of Applied Psychology, 73 (3), 421-35. Fischer, Agneta and Roger Giner-Sorolla (2016), "Contempt: Derogating Others While Keeping Calm," Emotion Review, 8 (4), 346-57. Fischer, Agneta H (2011), "Contempt: A Hot Feeling Hidden under a Cold Jacket," in Re-Constructing Emotional Spaces. From Experience to Regulation, ed. Radek Trnka, Karel Balcar and Martin Kuška, Prague, Czech Republic: College of Psychosocial Studies Press, 77-87. Fischer, Agneta H. (2018), "Why Hate Is Unique and Requires Others for Its Maintenance," Emotion Review, forthcoming. Fischer, Agneta H., Eran Halperin, Daphna Canetti, and Alba Jasmi (2018), "Why We Hate," Emotion Review, forthcoming. Fischer, Agneta H. and Ira J. Roseman (2007), "Beat Them or Ban Them: The Characteristics and Social Functions of Anger and Contempt," Journal of Personality and Social Psychology, 93 (1), 103-15. Fisher, Robert J. (1993), "Social Desirability Bias and the Validity of Indirect Questioning," Journal of Consumer Research, 20 (2), 303-15.   80 Fitness, Julie (2000), "Anger in the Workplace: An Emotion Script Approach to Anger Episodes between Workers and Their Superiors, Co‐Workers and Subordinates," Journal of Organizational Behavior, 21 (2), 147-62. Fitness, Julie and Garth J. O. Fletcher (1993), "Love, Hate, Anger, and Jealousy in Close Relationships: A Prototype and Cognitive Appraisal Analysis," Journal of Personality and Social Psychology, 65 (5), 942-58. Fournier, Susan (1998), "Consumers and Their Brands: Developing Relationship Theory in Consumer Research," Journal of Consumer Research, 24 (4), 343-53. ______ (2009), "Lessons Learned About Consumers’ Relationships with Brands," in Handbook of Brand Relationships, ed. Debbie MacInnis, C. Whan Park and Joseph W. Priester, New York: M.E. Sharpe, 5-23. Fournier, Susan and Claudio Alvarez (2013), "Relating Badly to Brands," Journal of Consumer Psychology, 23 (2), 253–64. Gander, Kashmira (2015), "Ukip Named Uk's Most Hated Brand, Followed by the Tories and Marmite," https://web.archive.org/web/20160706215946/http://www.independent.co.uk/news/uk/home-news/ukip-named-uks-most-hated-brand-followed-by-the-tories-and-marmite-10049886.html.   81 Gervais, Matthew M. and Daniel M. T. Fessler (2017), "On the Deep Structure of Social Affect: Attitudes, Emotions, Sentiments, and the Case of “Contempt”," Behavioral and Brain Sciences, 40, e225. Gollwitzer, Mario and Markus Denzler (2009), "What Makes Revenge Sweet: Seeing the Offender Suffer or Delivering a Message?," Journal of Experimental Social Psychology, 45 (4), 840-44. Goodwin, Geoffrey P., Jared Piazza, and Paul Rozin (2014), "Moral Character Predominates in Person Perception and Evaluation," Journal of Personality and Social Psychology, 106 (1), 148-68. Grégoire, Yany, Daniel Laufer, and Thomas M. Tripp (2010), "A Comprehensive Model of Customer Direct and Indirect Revenge: Understanding the Effects of Perceived Greed and Customer Power," Journal of the Academy of Marketing Science, 38 (6), 738-58. Grégoire, Yany, Thomas M. Tripp, and Renaud Legoux (2009), "When Customer Love Turns into Lasting Hate: The Effects of Relationship Strength and Time on Customer Revenge and Avoidance," Journal of Marketing, 73 (6), 18-32. Griskevicius, Vladas and Douglas T. Kenrick (2013), "Fundamental Motives: How Evolutionary Needs Influence Consumer Behavior," Journal of Consumer Psychology, 23 (3), 372-86. Haidt, Jonathan (2003), "The Moral Emotions," in Handbook of Affective Sciences, ed. Klaus Scherer, Richard Davidson and H. Hill Goldsmith, New York: Oxford University Press, 852-70.   82 Halperin, Eran (2008), "Group-Based Hatred in Intractable Conflict in Israel," Journal of Conflict Resolution, 52 (5), 713-36. ______ (2012), "Intergroup Hatred: Psychological Dimensions," in The Encyclopedia of Peace Psychology, ed. Daniel J. Christie, Chichester: Blackwell, 557-61. Halperin, Eran, Daphna Canetti, and Shaul Kimhi (2012), "In Love with Hatred: Rethinking the Role Hatred Plays in Shaping Political Behavior," Journal of Applied Social Psychology, 42 (9), 2231-56. Halperin, Eran, Alexandra G. Russell, Carol S. Dweck, and James J. Gross (2011), "Anger, Hatred, and the Quest for Peace: Anger Can Be Constructive in the Absence of Hatred," Journal of Conflict Resolution, 55 (2), 274-91. Harris, Christine R. (2018), "Understudied Negative Emotions: What They Can Tell Us About the Nature of Emotions," Emotion Review, forthcoming. Hayes, Andrew F. (2018), Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach, New York, NY: Guilford Press. He, Stephen X. and Samuel D. Bond (2015), "Why Is the Crowd Divided? Attribution for Dispersion in Online Word of Mouth," Journal of Consumer Research, 41, 1509-27. Henkel, Alexander P., Johannes Boegershausen, Joandrea Hoegg, Karl Aquino, and Jos Lemmink (2018), "Discounting Humanity: When Consumers Are Price Conscious Employees Appear Less Human," Journal of Consumer Psychology, 28 (2), 272-92.   83 Hoch, Stephen J. and John Deighton (1989), "Managing What Consumers Learn from Experience," Journal of Marketing, 53 (2), 1-20. Hogreve, Jens and Dwayne D. Gremler (2009), "Twenty Years of Service Guarantee Research: A Synthesis," Journal of Service Research, 11 (4), 322-43. Hutcherson, Cendri A. and James J. Gross (2011), "The Moral Emotions: A Social–Functionalist Account of Anger, Disgust, and Contempt," Journal of Personality and Social Psychology, 100 (4), 719-37. Isaac, Mathew S. and Kent Grayson (2017), "Beyond Skepticism: Can Accessing Persuasion Knowledge Bolster Credibility?," Journal of Consumer Research, 43 (6), 895-912. Janakiraman, Ramkumar, Catarina Sismeiro, and Shantanu Dutta (2009), "Perception Spillovers across Competing Brands: A Disaggregate Model of How and When," Journal of Marketing Research, 46 (4), 467-81. Johnson, Allison R., Maggie Matear, and Matthew Thomson (2011), "A Coal in the Heart: Self-Relevance as a Post-Exit Predictor of Consumer Anti-Brand Actions," Journal of Consumer Research, 38 (1), 108-25  Joireman, Jeff, Yany Grégoire, Berna Devezer, and Thomas M. Tripp (2013), "When Do Customers Offer Firms a “Second Chance” Following a Double Deviation? The Impact of Inferred Firm Motives on Customer Revenge and Reconciliation," Journal of Retailing, 89 (3), 315-37.   84 Kähr, Andrea, Bettina Nyffenegger, Harley Krohmer, and Wayne D. Hoyer (2016), "When Consumers Harm Your Brand: The Phenomenon of Consumer Brand Sabotage," Journal of Marketing, 80 (3), 25-41. Kenrick, Douglas T., Vladas Griskevicius, Steven L. Neuberg, and Mark Schaller (2010), "Renovating the Pyramid of Needs: Contemporary Extensions Built Upon Ancient Foundations," Perspectives on Psychological Science, 5 (3), 292-314. Kim, Kwang-Ho and Wenpin Tsai (2012), "Social Comparison among Competing Firms," Strategic Management Journal, 33 (2), 115-36. Kumar, Piyush (2005), "Brand Counterextensions: The Impact of Brand Extension Success Versus Failure," Journal of Marketing Research, 42 (2), 183-94. Landy, Justin F., Jared Piazza, and Geoffrey P. Goodwin (2016), "When It’s Bad to Be Friendly and Smart: The Desirability of Sociability and Competence Depends on Morality," Personality and Social Psychology Bulletin. Legere, John (2017), " T-Mobile’s Ceo on Winning Market Share by Trash-Talking Rivals," Harvard Business Review, 95 (1), 37–42. Li, Yexin Jessica, Douglas T. Kenrick, Vladas Griskevicius, and Steven L. Neuberg (2012), "Economic Decision Biases and Fundamental Motivations: How Mating and Self-Protection Alter Loss Aversion," Journal of Personality and Social Psychology, 102 (3), 550-61.   85 Lisjak, Monika and Angela Y. Lee (2014), "The Bright Side of Impulse: Depletion Heightens Self-Protective Behavior in the Face of Danger," Journal of Consumer Research, 41 (1), 55-70. Luo, Xueming, Michael Wiles, and Sascha Raithel (2013), "Make the Most of a Polarizing Brand," Harvard Business Review, 91 (11), 29-31. Matsumoto, David (2009), "Hatred," in The Cambridge Dictionary of Psychology, ed. David Matsumoto, Cambridge, UK: Cambridge University Press, 587. McCullough, Michael E., Robert Kurzban, and Benjamin A. Tabak (2013), "Revenge and Forgiveness - Cognitive Systems for Revenge and Forgiveness," Behavioral and Brain Sciences, 36 (1), 1–15. Mende, Martin, Ruth N. Bolton, and Mary Jo Bitner (2013), "Decoding Customer–Firm Relationships: How Attachment Styles Help Explain Customers' Preferences for Closeness, Repurchase Intentions, and Changes in Relationship Breadth," Journal of Marketing Research, 50 (1), 125-42. Mochon, Daniel (2013), "Single-Option Aversion," Journal of Consumer Research, 40 (3), 555-66. Nasr Bechwati, Nada and Maureen Morrin (2003), "Outraged Consumers: Getting Even at the Expense of Getting a Good Deal," Journal of Consumer Psychology, 13 (4), 440-53.   86 ______ (2007), "Understanding Voter Vengeance," Journal of Consumer Psychology, 17 (4), 277-91. Neel, Rebecca, Douglas T. Kenrick, Andrew Edward White, and Steven L. Neuberg (2016), "Individual Differences in Fundamental Social Motives," Journal of Personality and Social Psychology, 110 (6), 887-907. O’Reilly, Kelley, David Paper, and Sherry Marx (2012), "Demystifying Grounded Theory for Business Research," Organizational Research Methods, 15 (2), 247-62. Oliver, Richard L. (2010), Satisfaction: A Behavioral Perspective on the Consumer, Armork, New York; London: M. E. Sharp. Opotow, Susan and Sara I. McClelland (2007), "The Intensification of Hating: A Theory," Social Justice Research, 20 (1), 68-97. Oppenheimer, Daniel M., Tom Meyvis, and Nicolas Davidenko (2009), "Instructional Manipulation Checks: Detecting Satisficing to Increase Statistical Power," Journal of Experimental Social Psychology, 45 (4), 867-72. Paharia, Neeru, Jill Avery, and Anat Keinan (2014), "Positioning Brands against Large Competitors to Increase Sales," Journal of Marketing Research, 51 (6), 647-56. Park, Whan C., Deborah J. MacInnis, Joseph Priester, Andreas B. Eisingerich, and Dawn Iacobucci (2010), "Brand Attachment and Brand Attitude Strength: Conceptual and   87 Empirical Differentiation of Two Critical Brand Equity Drivers," Journal of Marketing, 74 (6), 1-17. Peer, Eyal, Laura Brandimarte, Sonam Samat, and Alessandro Acquisti (2017), "Beyond the Turk: Alternative Platforms for Crowdsourcing Behavioral Research," Journal of Experimental Social Psychology, 70, 153-63. Piazza, Jared, Geoffrey P. Goodwin, Paul Rozin, and Edward B. Royzman (2014), "When a Virtue Is Not a Virtue: Conditional Virtues in Moral Evaluation," Social Cognition, 32 (6), 528-58. Rempel, John K. and Christopher T. Burris (2005), "Let Me Count the Ways: An Integrative Theory of Love and Hate," Personal Relationships, 12 (2), 297-313. Rempel, John K., Christopher T. Burris, and Darius Fathi (2018), "Hate: Evidence for a Motivational Conceptualization," Motivation and Emotion, forthcoming. Roehm, Michelle L. and Alice M. Tybout (2006), "When Will a Brand Scandal Spill over, and How Should Competitors Respond?," Journal of Marketing Research, 43 (3), 366-73. Romani, Simona, Silvia Grappi, and Daniele Dalli (2012), "Emotions That Drive Consumers Away from Brands: Measuring Negative Emotions toward Brands and Their Behavioral Effects," International Journal of Research in Marketing, 29 (1), 55-67. Roseman, Ira J. and Amanda K. Steele (2018), "Schadenfreude, Gluckschmerz, Jealousy, and Hate—What (and When, and Why) Are the Emotions?," Emotion Review, forthcoming.   88 Royzman, Edward B., Clark McCauley, and Paul Rozin (2005), "From Plato to Putnam: Four Ways to Think About Hate," in The Psychology of Hate, ed. Robert J. Sternberg, Washington, DC: American Psychological Association, 3-35. Rozenkrants, Bella, S. Christian Wheeler, and Baba Shiv (2017), "Self-Expression Cues in Product Rating Distributions: When People Prefer Polarizing Products," Journal of Consumer Research, 44 (4), 759-77. Sengupta, Jaideep, Darren W. Dahl, and Gerald J. Gorn (2002), "Misrepresentation in the Consumer Context," Journal of Consumer Psychology, 12 (2), 69-79. Sinaceur, Marwan and Larissa Z. Tiedens (2006), "Get Mad and Get More Than Even: When and Why Anger Expression Is Effective in Negotiations," Journal of Experimental Social Psychology, 42 (3), 314-22. Spencer, Steven J., Mark P. Zanna, and Geoffrey T. Fong (2005), "Establishing a Causal Chain: Why Experiments Are Often More Effective Than Mediational Analyses in Examining Psychological Processes," Journal of Personality and Social Psychology, 89 (6), 845-51. Spiggle, Susan (1994), "Analysis and Interpretation of Qualitative Data in Consumer Research," Journal of Consumer Research, 21 (3), 491-503. Spirit (2014), "State of the Hate Report," http://tiny.cc/spirit_hate.   89 Sprott, David, Sandor Czellar, and Eric Spangenberg (2009), "The Importance of a General Measure of Brand Engagement on Market Behavior: Development and Validation of a Scale," Journal of Marketing Research, 46 (1), 92-104. Stebbins, Samuel , Evan Comen, Michael B. Sauter, and Charles Stockdale (2018), "Bad Reputation: America’s Top 20 Most-Hated Companies," http://tiny.cc/usa_today. Steegen, Sara, Francis Tuerlinckx, Andrew Gelman, and Wolf Vanpaemel (2016), "Increasing Transparency through a Multiverse Analysis," Perspectives on Psychological Science, 11 (5), 702-12. Steffel, Mary and Elanor F. Williams (2018), "Delegating Decisions: Recruiting Others to Make Choices We Might Regret," Journal of Consumer Research, 44 (5), 1015-32. Sternberg, Robert J. (2003), "A Duplex Theory of Hate: Development and Application to Terrorism, Massacres, and Genocide," Review of General Psychology, 7 (3), 299-328. ______ (2005), "Understanding and Combating Hate," in The Psychology of Hate, ed. Robert J. Sternberg, Washington, DC: American Psychological Association, 37-49. Sternberg, Robert J. and Karin Sternberg (2008), The Nature of Hate, New York: Cambridge University Press. Sundie, Jill M., Douglas T. Kenrick, Vladas Griskevicius, Joshua M. Tybur, Kathleen D. Vohs, and Daniel J. Beal (2011), "Peacocks, Porsches, and Thorstein Veblen: Conspicuous   90 Consumption as a Sexual Signaling System," Journal of Personality and Social Psychology, 100 (4), 664-80. Tausczik, Yla R. and James W. Pennebaker (2010), "The Psychological Meaning of Words: Liwc and Computerized Text Analysis Methods," Journal of Language and Social Psychology, 29 (1), 24-54. Thompson, Craig J., William B. Locander, and Howard R. Pollio (1990), "The Lived Meaning of Free Choice: An Existential-Phenomenological Description of Everyday Consumer Experiences of Contemporary Married Women," Journal of Consumer Research, 17 (3), 346-61. Thomson, Matthew, Deborah J. MacInnis, and Whan C. Park (2005), "The Ties That Bind: Measuring the Strength of Consumers’ Emotional Attachments to Brands," Journal of Consumer Psychology, 15 (1), 77-91. van Doorn, Janne (2018), "Anger, Feelings of Revenge, and Hate," Emotion Review, 1754073918783260. van Kleef, Gerben A., Carsten K. W. De Dreu, and Antony S. R. Manstead (2004), "The Interpersonal Effects of Anger and Happiness in Negotiations," Journal of Personality and Social Psychology, 86 (1), 57-76. Wang, Feng, Xuefeng Liu, and Eric Fang (2015), "User Reviews Variance, Critic Reviews Variance, and Product Sales: An Exploration of Customer Breadth and Depth Effects," Journal of Retailing, 91 (3), 372-89.   91 Westbrook, Robert A. (1987), "Product/Consumption-Based Affective Responses and Postpurchase Processes," Journal of Marketing Research, 24 (3), 258-70. Winterich, Karen Page, Vikas Mittal, and Andrea C. Morales (2014), "Protect Thyself: How Affective Self-Protection Increases Self-Interested, Unethical Behavior," Organizational Behavior and Human Decision Processes, 125 (2), 151-61. Yang, Linyun W. and Pankaj Aggarwal (2018), "No Small Matter: How Company Size Affects Consumer Expectations and Evaluations," Journal of Consumer Research, forthcoming. Yim, Chi Kin (Bennett), David K. Tse, and Kimmy Wa Chan (2008), "Strengthening Customer Loyalty through Intimacy and Passion: Roles of Customer–Firm Affection and Customer–Staff Relationships in Services," Journal of Marketing Research, 45 (6), 741-56. Zarantonello, Lia, Simona Romani, Silvia Grappi, and Richard P. Bagozzi (2016), "Brand Hate," Journal of Product & Brand Management, 25 (1), 11-25. Zeki, Semir and John Paul Romaya (2008), "Neural Correlates of Hate," PLOS ONE, 3 (10), e3556.      92 Appendices  A1. Brand Hatred Prevalence Study Participants responded to the following question: Are there any brands that you hate? (“Yes” vs. “No”). If they responded with “Yes”, they were prompted to list up to five brands they hated with one line each per brand. The percentages of the categories presented in the introduction are based on all brands (N = 124) that were listed at least three times (after correcting for spelling errors). The most hated brands were Wal-Mart (4.8%), Apple (4.1%), Nike (1.9%), Nestle (1.8%), and Pepsi (1.8%). The most hated categories in order of prevalence were: apparel (18%), supermarket and service (18%), fast-moving consumer goods (12%), and technology (10%).     93 A2. Interview Outline (Qualitative Study) Research agenda Interviewer statements and questions Introduction to research project Purpose of the project Thank you very much for participating in our interview today.   Before we start, we would like to kindly ask you to read and sign our confidentiality agreement.  Is it ok for us to record the interview from here on?  We are trying to understand your experience with different connections and feelings you have about different consumer brands.   Before we start, we just want to make it clear that there are really no right or wrong answers to any of our questions, so just really feel free to openly discuss with us whatever comes to your mind. We are interested in strong opinions, so please do not hold back. Sometimes, it can be difficult to remember things that happened in the past, so just take your time. There is really no hurry today. And if you feel that a question is unclear, please just say so and we will try to rephrase and tackle it differently.   Practical information about interview The interview will take not longer than 60 minutes. This interview will be kept strictly confidential and will be available only to members of the research team. Excerpts of this interview may be made part of the final research report, but under no circumstances will your name or identifying characteristics be included in this report.  Making respondent comfortable So maybe we can start with us introducing ourselves and then perhaps you could tell us more about yourself…     94 Topics of the actual interview  Establishing mutual understanding of brand Before we delve deeper into the topic of today, we would like to make sure that we all understand what brands are so we are talking about the same thing. What do you think about when you hear the term “brand”?   So if you could give us some examples of what you consider a brand, we can take it from there.  Today, we are specifically interested in brands you have a very negative connections.  Components of negative consumer-brand connection   Option A: How did this experience you just described make you feel at the time?  So you say that it makes you feel …, is there more to it? / Does it also make you feel other ways?  PROBE FOR DISSATISFACTION/DISLIKE  Option B: You said you were XXX with the YY. Can you tell me more about this XX? Back-up option: Can you elaborate further on this negative connection with brand X? What does it consist of?  Consequences of negative consumer-brand connection   Thank you for sharing your experience, we would now like to move on and focus on the effects of your feelings toward XX brand in a broader sense.  What other behaviors/consequences do you possibly take towards this brand in the future?   We would also like to know whether and how your feelings toward XX brand affect your behavior in general?  How do your feelings impact your behavior toward other brands than THIS FOCAL ONE?  PROBE: Hmmm…Can you tell me a bit more on why you say that?   95 PROBE: What if you could create YOUR OWN NEW X BRAND, what would it look/feel/be like? Category How do you perceive other brands in the YY-category?  If you shop for an YY-category, how does your feeling toward XX affect how you approach your decision?  What do you focus on when shopping in YY-category?  PROBE: You said you would choose YXZ type of service/product? Why? [probe for self-protection vs. revenge]  Could you imagine ever purchasing from this brand again?  If yes: What would it take for you to repurchase the brand you currently feel negative about? If no: Why not?     Triangulate hatred vs. dislike vs. dissatisfaction:  Before you said you … X brand, are there any other brands you have similar feelings for? Can you think of any brand that you …      … dislike?      … are dissatisfied with?  How similar is your reaction to that brand to your reaction to X we talked about at the beginning of the interview.  Wrapping up the interview Other factors Are there any other factors that cause you to feel the way you do abut brand X that we have not discussed so far?  Is there anything else you would like to add? Is there anything I have not asked regarding your experiences that you‘d like to tell me? Do you feel the same way to any other brand? How would you describe your feelings toward XX brand in one word? [PROBE for dissatisfaction]    96 Provide contact information for future questions, feedback or additional insights. This was our last question. Thank you again for your time today. We very much appreciate your insights.   If there is anything else that comes to your mind regarding this research later on, please do not hesitate to send us an email. References enquiry  If you know someone, a friend or relative, who also feels negatively towards a brand, please feel free to forward our contact details and they could also participate. However, we ask you not to talk about the content of the interview.         97 A3. Overview of Informants (Qualitative Study)  # Name* Age range Ethnicity Education Occupation Household Income Hated Brands Illustrative themes and structure of sentiment 1 Martha 30-39 Caucasian College Education Consultant $75,000 to less than $100,000 Fido, Sears, Under Armour Martha felt exploited and disrespected by her mobile telephone provider, Fido. This feeling was aggravated by a deep sense of betrayal after having been a loyal customer for more than a decade. She expressed deep-seated frustration and anger with the brand. As a reaction of her experiences she continues to engage in negative word-of-mouth about the brand. One of the most important aspects when searching for a new provider was to find a firm that would not wrong her and she was dismayed with the category structure in the market (i.e., the absence of a fundamentally different category with reliable providers).  2 Heather 30-39 European Caribbean Ph.D. Research Support and Administrator $75,000 to less than $100,000 Walmart, Gap, Monsanto,  Coca-Cola Heather’s account of her hatred for Walmart centered around the damage she perceives the retailer exerts on communities that are important to her. Her sentiment for Walmart was fueled by indirect consumption of news about the brand and well as her despair about the effect of the opening of a Walmart in the town where her parents live. She tried to contact the hated brands (e.g., Gap) to enact change from them. Her hatred for Gap and the resultant avoidance of the brand meant significant costs to her in so far given her reliance on the brand as one of the few large size cloth providers in the market. While her hatred for Walmart as well as for the other hated brands was characterized by feeling powerless and exploited, she also expressed contempt for the people responsible for managing these brands driven largely by their neglect for consumer welfare. When shopping in these categories, she explicitly tried to search for different providers. Her shopping behavior was focused on taking care of herself rather than getting back at the hated brands. She actively tracked the fortunes of the hated brand and rejoiced at seeing Gap struggling in the marketplace.  3 Tania 50-59 Asian College Program Manager $100,000 or more Cactus Club Tania’s sentiment toward Cactus Club was triggered by the presence of a painting in the servicescape that she found extremely offensive. She was disgusted by the exposure to this artwork. However, more critical to development of her sentiment for the brand was the reaction of the firm. She felt not appreciated and somewhat humiliated by the management response that this painting had been picked by an interior designer. She vowed to never return to this chain of restaurants. After this experience, she became less interested in dining in chain restaurants and showed greater inclinations to dine at local single-location eateries.                 98 # Name* Age range Ethnicity Education Occupation Household Income Hated Brands Illustrative themes and structure of sentiment 4 Lucy 30-39 Asian College Art educator $100,000  or more West Coast Kids, British Airways Lucy came to a hate a local baby store after several store employees gave her inaccurate and misleading product information that if followed would have let her to purchase unnecessarily expensive products for her toddler. She attributed this behavior at least partially to some ulterior profit-maximizing corporate motives. As a consequence of her hatred-evoking encounters she significantly reduced her engagement with the brand. In light of the limited number of competitors it was not possible to fully avoid the store. She would, however, no longer browse the aisles for visiting the store without a clear purpose and did more of her baby shopping online. Moreover, she did more careful prior research online in order to avoid at the mercy of ill-meaning store employees.   5 Barry 40-49 European College Server $75,000  to less than $100,000 Aldo Barry’s hatred for the shoe retailer Aldo resulted from a severe product failure shortly after the purchase. This experience was particularly intense for him since he was new to and unfamiliar with the local marketplace. Barry expressed feelings of helplessness resulting from being a newcomer to the country with limited social resources (e.g., friends that could provide recommendations). He recalled the experience of anger and rage at the brand. Barry continues to engage in negative word-of-mouth about the brand. Subsequently, for the next purchase in the category he purposefully select a very different kind of store that had larger assortments featuring multiple brands. Even years after the initial brand experience, Barry still remained bewildered when seeing Aldo survive and thrive in the marketplace. His hatred for Aldo led to a more generalized concern about others taking advantage of him in the marketplace beyond his actions as a consumer (e.g., deciding on whether to marry his current girlfriend or not).  6 Mark 30-39 Caucasian Master’s Financial Analyst $100,000 or more Loblaws Mark came to hate Loblaws because of its corporate pricing strategies. In his view, the brand was deliberately trying to overprice relatively healthy items to lure people into purchasing their own private label processed food items. He expressed anger, outrage, and contempt for these managerial practices. This practice reduced the possibility for one-stop shopping for him and he deliberately tried to avoid any stores associated with this brand. For grocery shopping he purposefully sought brands from different categories (e.g., WholeFoods). When scanning the marketplace, he acted in a relatively risk averse manner by grouping brands falling in the between categories (e.g., a local supermarket chain) with the category containing Loblaws.   7 Anna 40-49 Caucasian Master’s Teacher $75,000  to less than $100,000 American Apparel Exposure to the advertisements in the storefront of American Apparel triggered strong emotional reactions involving disgust and intense anger from Anna. She found the manner in which the younger models – in particular the female models – were presented deeply offensive. Her indignation for the brand was amplified by the lack of response after writing a complaint letter to the firm; she expressed contempt for the management of the firm. Her sentiments toward American Apparel let her to scrutinize the marketplace more carefully and she mentioned how it led her to feel safer at more local brands such as The Bay or Roots.   99 # Name* Age range Ethnicity Education Occupation Household Income Hated Brands Illustrative themes and structure of sentiment 8 Susan 30-39 Asian College Auditor $75,000  to less than $100,000 Tesla, Apple, The Gap Susan’s accounts of hatred for the different brands were characterized by a disdain for other consumers who would blindly follow the promise of brands such as Apple or Tesla. She expressed concern about the practices of these brands and reported outrage at Tesla’s history of accidents. Based on her sentiments, she sought car brands that do not offer electric cars at all. Similarly, she felt that Apple was overly restrictive on consumer choice and was committed to choosing phones that would allow here greater freedom of choice.  9 Paul 30-39 South Asian Master’s Teaching Assistant $30,000  to less than $50,000 PETA Paul found the tactics used in the advertisements by PETA deeply offensive and counterproductive for a cause he felt strong personal connection to, the protection of animals. He expressed anger and frustration at the organization’s constant push for attention at all cost. While he perceived the tactics to be specific to PETA, it made him approach other non-governmental organization more critically and he was more likely to perceive them as compromised.   10 Will 40-49 Asian College Dentist $75,000  to less than $100,000 None Will expressed dislike for a number of brands including Range Rover and Apple. However, unlike consumers experiencing hatred, his sentiments toward these brands were less likely to be characterized by strong negative emotions, but rather by disappointment about the price-to-quality ratio, overpriced products, or a lack of consistency between the brand values and service experiences.  11 Gareth 30-39 South Asian Master’s Project Manager $100,000 or more None Gareth expressed disdain for a variety of brands including Lululemon and Coach. For instance, his sentiment toward Lululemon was void of strong negative emotions, but rather centered on a sense of irritation about the brand’s commercialization of yoga for their purposes. He felt that the brand was deliberately designed to be snobbish. Gareth also felt somewhat upset by the comments made by the previous CEO of Lululemon involving some degree of body shaming of certain consumers.   12 Nick 40-49 African College Gardener $30,000  to less than $50,000 None Nick expressed opposition to large corporations like Nestlé and Coca-Cola. His socio-political perspective primarily drove his feelings toward these brands on the influence of these large corporations on communities. He expressed dissatisfaction with how these companies treat their workers. However, his responses did not include stronger negative emotional reactions. His way of dealing with his feelings for these brands was to purposefully avoid any of their products in line with his lifestyle oriented toward living a healthy life.  Notes: The names used in the table above are pseudonyms. All interviews were conducted at a neighborhood café of the respondent’s choosing.   100 A4. Cuisine as a Category Marker  (Study 1)  We gave respondents a list of five potential aspects of restaurants modeled after review sites such as Yelp and TripAdvisor (i.e., type of cuisine, price level, location, restaurant style, and parking). The results showed that the majority of consumers selected type of cuisine as their first consideration when choosing a restaurant (N = 73, 70.9%). This was significantly greater than chance (χ2(1) = 17.95, p < .001). Of those participants who did not list cuisine as the most important (N = 30), 56.7% listed cuisine as the second and 23.3% as the third most important aspect when considering where to dine out. Only six respondents (5.8%) did not list the type of cuisine as a top three criteria. Accordingly, type of cuisine constitutes a commonly applied criterion for restaurant choice and, hence, was used as the categorization variable in the first study in my dissertation.     101 A5. Review- and Reviewer-Specific Control Variables (Study 1) Variable Rationale Operationalization Review-specific controls   review age writers of older reviews had more time to visit and review other same-subcategory restaurants number of days between the posting data of the review and the date of date of data collection number of words writers of longer reviews might be more engaged in reviewing on Yelp and thus have a higher propensity to review other same-subcategory restaurants word count of review text restaurant dummies there might be differences with respect to the propensity to visit restaurants in a category unrelated to the focal experience dummy variables for the Indian, Italian, Japanese, and Thai restaurant with the Chinese restaurant serving as the baseline Reviewer-specific controls   number of reviews before users with more reviews prior to the focal review might be more engaged on Yelp count of all reviews written prior to the focal review number of reviews after users who are more active on Yelp after the focal review have a higher propensity to review another same-subcategory restaurant count of all reviews written after the focal review friends more friends could signal engagement on Yelp and a higher likelihood to post reviews number of friends on Yelp    102 A6. Robustness Checks (Study 1)   To ensure that the effects were driven by feelings about a particular brand (e.g., a specific Indian restaurant) rather than hatred of the whole category, I used several ways of controlling for a reviewer’s prior involvement and feelings toward the category. Specifically, I used three ways to control for consumer engagement with the subcategory. First, I dummy coded whether or not the Yelp user had previously reviewed a restaurant in the subcategory (0 = first review of a restaurant in the subcategory, 1 = previously reviewed a restaurant from the subcategory) and interacted this dummy variable with the hatred dummy. In this model the effect of the focal hatred dummy main effect remained significant after adding these controls both for the binary subcategory return (βexp = -.47, SE = .19, Wald χ2(1) = 6.10, p = .013) as well as the continuous inter-review time (b = 194.90, SE = 63.02, t(1079) = 3.09, p = .002). Neither the main effect of the first-time dummy (both ps >. 07) nor its interaction with the hatred dummy were significant (both ps > .15). Second, I created a control variable for consumers’ category preference as the share of same-subcategory visits of total reviews prior to the critical visit (M = .05, SD = .12, range 0-1). Using this control, the effect of hatred dummy was significant both the binary subcategory return (βexp = -.41, SE = .17, Wald χ2(1) = 5.85, p = .016) and significant for the continuous inter-review time (b = 179.70, SE = 56.06, t(1079) = 3.21, p < .01). The main effect of the share of prior same-subcategory variable was marginally significant for the binary control variable (βexp = 1.58, SE = .84, Wald χ2(1) = 3.58, p = .058), but not the continuous inter-review time (p > .22). The interaction between this variable and the hatred dummy was not significant in any of the models (both ps > .23). Third, I ran another models accounting for prior valence of same-subcategory reviews using two dummies (i.e., negative = less than 3 stars average ratings prior to the focal experience, M = .07, SD = .26; neutral to positive = 3 stars or higher average   103 ratings prior to the focal experience, M = 0.29, SD = 0.46) with no prior ratings serving as the baseline. In this model, I again I found a significant effect of the focal hatred dummy on the binary subcategory return (βexp = -.46, SE = .19, Wald χ2(1) = 6.07, p = .014) as well as the continuous inter-review time (b = 195.41, SE = 62.96, t(1077) = 3.10, p = .002). In contrast to the prior models, I found some evidence that prior category experiences might buffer against the effects of hatred as the interaction between the hatred dummy and the prior positive category reviews was marginally significant in both models (binary: βexp = -.61, SE = .35, Wald χ2(1) = 3.05, p = .081; continuous: b = -221.52, SE = 116.55, t(1077) = 1.90, p = .058). In contrast, none of the main effects or interactions of the negative dummy was significant (all ps > .30).  I also used the net negativity as expressed in the review content as function of the use of negative emotion words minus positive emotion words (as a percentage of total word count) instead of the star ratings as proxies for hatred. Binary logistic regressions with controls revealed a significant negative effect of the net negativity of a review (βexp = -.05, SE = .02, Wald χ2(1) = 6.57, p = .01) on subcategory return. In addition, the effect of the net negativity on the inter-review time (b = 9.56, SE = 5.78, t(1081) = 1.65, p = .099) was marginally significant.  104 A7. Multiverse Analysis (Study 1)      105 A8. Stimuli and Consumption Scenarios (Study 2) Hatred: drag-and-drop task  Below you will find a list of several brands from different categories listed under Items. Please drag-n-drop each brand into the box that best describes how you feel toward this brand. If you hate a brand drag it into the "I hate this brand" box, if you do not hate a particular brand please drag it into the "I do not hate this brand" box.  Items Ryanair Starbucks Volkswagen BP  I hate this brand I do not hate this brand      Consumption Scenarios The following four consumption scenarios were presented to participants. Their order was counterbalanced.   Coffee Imagine you were about to go to a coffee shop on your way to the office. While strolling along the street, you see a Costa Coffee store and a local coffee shop called BelCafé. Please indicate which of these two stores you would prefer...  (1) Definitely the BelCafé – (9) Definitely the Costa Coffee        106 Airlines Imagine you were about to book a flight. While searching for flights, you find that both British Airways and easyJet offer a flight to your destination. Please indicate which of these two airlines you would prefer... (1) Definitely British Airways – (9) Definitely easyJet  Gas station Please imagine you have to refuel. While driving, you see two petrol stations on your side of the road, a Shell station and Manor Service, a local station. Please indicate which of these two petrol stations you would prefer... (1) Definitely the Manor Service station– (9) Definitely the Shell station  Car rental Please imagine you are about to rent a car. As you study the website of the rental car company, you find that you can either rent a Lexus or a Mercedes. Please indicate which of these two cars you would prefer to rent...  (1) Definitely the Lexus – (9) Definitely the Mercedes        107 A9. Tax Advisor Scenario and Dependent Variable (Study 3) Introduction Please imagine that tax season is about to begin. When thinking about filing your taxes for this year, you reflect on your experiences with your tax advisor, George Adams, in the previous year. George Adams founded his own tax filling consultancy service firm four years ago. He offers consulting services on a variety of tax topics and assists his clients in negotiating with the IRS and other government agencies.  Hatred condition Last year I used George Adams to help me file my taxes. The initial meeting with George was promising. He seemed knowledgeable and has an easy-going personality. However, after the initial meeting, things did not go as expected. After gathering all my information and documents during our initial meeting, George told me that the taxes would be done in a week.  After a week passed, he did not contact me. I tried calling him multiple times, but it was not possible to get hold of him. When I finally reached his secretary, she was extremely rude and dismissive. I had to wait for more than 20 minutes before she eventually managed to put me through to him. George was extremely uncaring and brushed me off me off with an absurd excuse. He said he would call back later and just hung up on me. I was appalled and irritated.  Instead of calling back, it took George another six days before I finally got an e-mail from him informing me that my taxes were ready. He referred me to his secretary to schedule the appointment. She was extremely slow to respond and could only book me in two weeks later. When we finally sat down to go over my tax return, I realized that George had made several significant errors. If I had not caught them it would have surely gotten me into major trouble     108 with the IRS. Instead of apologizing, he blamed me for not providing the information, which was a blatant lie. I was disgusted by his behavior and full of contempt for him.  Two weeks later I received the final bill from his office, which was almost twice than what we had discussed. His greed and dishonesty were infuriating. Overall, I feel that I really hate George Adams.  Dissatisfaction condition Last year I used George Adams to help me file my taxes. The initial meeting with George was promising. He seemed knowledgeable and has an easy-going personality. However, after the initial meeting, things did not go as expected. After gathering all my information and documents during our initial meeting, George told me that the taxes would be done in a week.  After a week passed, he did not contact me. I had to call his office to get hold of him. When I reached his secretary, she was friendly but I had to wait for ten minutes before she managed to put me through to him. George apologized and told me he was currently with another client. He said he would get back to me. This was not what I expected and I was quite disappointed. It took George two more days before I finally got an e-mail from him informing me that my taxes were ready. He referred me to his secretary to schedule the appointment. She could only book me in at the end of the next week. When we finally sat down to go over my tax return, I realized George had made a few minor errors. If I had not caught them it’s possible there might have been an issue. He was apologetic and corrected the mistakes promptly. Still, I was rather unhappy with his service. Two weeks later I received the final bill from his office, which was more or less like we had discussed. Overall, I do not hate George Adams, but I was dissatisfied with him.      109 Indifference condition Last year I used George Adams to help me file my taxes. The initial meeting with George was promising. He seemed knowledgeable and has an easy-going personality.  After the initial meeting, things went more or less as expected. After gathering all my information and documents during our initial meeting, George told me that the taxes would be done in a week.  After a week passed, he did not contact me so I called his office. When I reached his secretary, she was friendly and helpful. She managed to put me through to him quickly. George was courteous, but told me that he was currently with another client. He said he would call back later.  George called me back later that day to inform me that my taxes were ready. He scheduled an appointment to go over my return later that week. When we sat down to go over my tax return, I realized that George had made one minor error. He was apologetic and corrected the mistake promptly. I was somewhat satisfied with the process. Two weeks later I received the final bill from his office, which was as we had discussed. Overall, I feel indifferent about George Adams.         110 Dependent variable: provider choice (choice options introduction) While still reflecting on your experiences with your tax advisor, George Adams, in the previous year, you begin to look for tax advisors for this year's taxes on the web.  Your search produces the following four options: - Kevin Price, CPA, CFP (tax consultant) - Shane Mason, CPA, CFP (tax consultant) - the Main Street branch of H&R Block® - the Broadway branch of Liberty Tax Service®  Choice Please select whom you would use to help you file your taxes. ! Kevin Price, CPA, CFP (tax consultant)  (1)  ! Shane Mason, CPA, CFP (tax consultant)  (2)  ! the Main Street branch of H&R Block®  (3)  ! the Broadway branch of Liberty Tax Service®  (4)          111 A10. Tax Advisor Pretest (Study 3) During the pretest, participants (N = 94) read one of the three scenarios presented above about their experiences with George Adams. Subsequently, they indicated to what extent this experience would make them feel each of the 17 emotions adapted from Romani et al. (2012) listed in alphabetical order. Based on prior research I created three factors from this list of emotions:  Hatred-related (α = .97):  Angry, Appalled, Disdainful, Disgusted, Enraged, Mad, Offended, Scornful  Dissatisfaction-related (α = .94):  Disappointed, Discontented, Dissatisfied  General negative affect (α = .90):  Annoyed, Anxious, Heartbroken, Sad, Unhappy, Upset   Emotional reactions to tax advisor scenario  Indifference Dissatisfaction Hatred Hatred-related 2.06 (1.46) a 3.59 (1.37) b 5.41 (1.37) c Dissatisfaction-related 2.81 (1.48) a 5.76 (.97) b 6.09 (1.41) b General negative affect 2.43 (1.43) a 4.01 (.99) b 4.39 (1.29) b Notes:  The table shows the mean and standard deviation (in parenthesis) for each emotion factor. Means with a  different superscript are significantly different from the other means in the row at p < .001 in post hoc tests  with Tukey’s correction.        112 A11. Hatred Manipulation Check (Studies 3 – 7) Across studies 3-7, I used the following fourteen items adopted from the Triangular Hatred Scale (Sternberg and Sternberg 2008) substituting “Brand Name” for the respective brand or provider name in each study. • I think that Brand Name is truly disgusting. • I would never knowingly associate with Brand Name. • I feel that one cannot trust Brand Name at all. • I have no sympathy whatsoever for Brand Name. • I can sometimes feel my heart beat faster from the rage I feel when I start thinking about Brand Name. • I feel intense anger when I think of Brand Name. • When I think of Brand Name I become very angry. • People need to commit themselves to the fight against brands/people* like Brand Name.  • People need to take an active role in speaking out against brands/people* like Brand Name.  • I am committed to the fight against brands/people* like Brand Name.  • We must never waiver in our fight against brands/people* like Brand Name.  • I cannot imagine that Brand Name will ever change his harmful behavior. • I would join a movement that is aimed at fighting against Brand Name. • The public should be informed comprehensively about the danger of Brand Name.   *  In all studies apart from study 2 where the focal provider was an individual, I used  “brands” rather than “people”.       113 A12. Hotel Categorization Pretest (Study 4a) Participants responded to the following question:  Below you will find a list of four midrange hotels in San Francisco, please use the drag-and-drop tool to arrange the hotels into categories.  Abri Hotel *** King George Hotel *** Best Western *** Holiday Inn ***   Category 1  Category 2           114 A13. Hotel Scenario (Studies 4-6) Category Manipulation  Please imagine the following scenario as vividly as possible. You have to travel to a mid-size American city multiple times per year. Some of these trips are for business purposes others are for private occasions. When going to this city you like to stay in an area called the Waterside. The last time you went to this city, you stayed at the WATERSIDE Inn – An Independent Hotel / WATERSIDE Inn – A member of GC Hotels & Resorts. You selected this hotel, a ***-establishment, because of its reviews on booking.com.           /     115 Feelings Toward the Focal Brand Manipulation Indifference Hatred In line with the reviews you read, your stay at the Waterside Inn [with category identifier] was reasonable. When you arrived you did not have to wait when checking in.   The employee you were dealing with was friendly and competent. Your room was clean and in good shape.  In line with your expectations, there was no fee for internet usage.   Overall you were somewhat satisfied with your stay at Waterside Inn [with category identifier].   When you think about Waterside Inn [with category identifier], you feel indifferent.        In contrast to the reviews you read, your stay at Waterside Inn [with category identifier] was incredibly horrible. When you arrived you had to wait for 1 hour when checking in.   The employee you were dealing with was rude and hostile. Your room was very dirty and in terrible shape.  Completely defying your expectations, there was a huge fee for internet usage ($25 per day).  Overall you were extremely dissatisfied with your stay at Waterside Inn [with category identifier].   When you think about Waterside Inn [with category identifier], you feel extremely angry, contemptuous and disgusted.  Category identifier = “An Independent Hotel“ or “A member of GC Hotels & Resorts"     116 A14. Overview of the Consumers’ Choice Set (Studies 4a, 4b, & 6)          117 A15. Brand Recall Question (Study 4a) Please recall the story about a hotel stay that you were asked to imagine at the very beginning of this study and select the correct response. The WATERSIDE Inn is … - an independent hotel  - a member of GC Hotels & Resorts - a member of the LL Hotel Group  - don't remember       118 A16. Competitor Strategies (Study 6)   Equal Quality Reputation Condition  Superior Quality Reputation Condition  Superior Quality Reputation and Money-Back Guarantee Condition      119 A17. Consumers’ Self-Protection Concerns (Study 6)  To further bolster my self-protection account, I employed a different measure of these concerns that was more specific to the choice context. Specifically, participants indicated their perceptions of the relative safety of the two choice options on the following seven-point scale: "Which of these two hotels is the safer option to pick in order to minimize the possibility of experiencing unpleasant or negative events during your stay?" (1 = Definitely the Downtown Hotel; A member of the LL Hotel Group, 4 = Equal likelihood of negative experiences, 7 = Definitely the Quality Hotel; An Independent Hotel). To examine whether the money-back guarantee was effective in addressing consumer self-protection concerns (and thereby blocking the potential preference shifts induced by brand hatred), I conducted a moderated mediation analysis. In the moderated mediation analysis, feelings toward the focal brand acted as the independent variable (0 = indifference, 1 = hatred) and the quality advantage plus the money-back guarantee as the moderator (0 = absent, 1 = present). I also controlled for the quality advantage of the same-subcategory brand (0 = equal, 1 = superior).  In contrast to studies 4a and 4b, the within-choice-set measure of self-protection concerns related to the relative safety of the options acted as the mediator and the preference difference score which I obtained by subtracting the preference for independent hotel from the chain hotel preference was the dependent measure (PROCESS model 8; Hayes 2018). While the index of moderated mediation was only marginally significant with a 90% confidence interval from 5,000 bootstrap samples that does not include zero (.71, SE = .38, 90% CI: .08 to 1.33), the path coefficients and focal conditional indirect effects are fully consistent with my predictions. Specifically, I found that hatred led participants to perceive the brand from the other subcategory as safer (b = .81, SE = .23, t(324) = 3.59, p < .001), whereas a quality advantage plus the money-    120 back guarantee led participants to perceive the same-subcategory brand as safer (b = -.64, SE = .29, t(324) = 2.21, p = .028). Importantly, the interaction was marginally significant (b = -.77, SE = .39, t(324) = 1.97, p = .05), suggesting that quality advantage plus money-back guarantees was more important when participants hated the same-subcategory competitor. The main effect of a higher quality rating for the same-subcategory brand was not significant (b = -.28, SE = .22, t(324) = 1.24, p = .22). Consumers’ perceptions of relative option safety in turn predicted their preferences (b = -.93, SE = .07, t(323) = 12.50, p < .001). Accordingly, the conditional indirect effect of hatred via relative option safety on preferences was significant and negative when the same-subcategory brand did not have a quality advantage plus a money-back guarantee (a × b = -.75, SE = .21, 95% CI: -1.18 to -.35). In contrast, when the higher quality same-subcategory rival also offered a money-back guarantee, the conditional indirect effect of hatred via relative option safety was close to zero and non-significant (a × b = -.03, SE = .33, 95% CI: -.69 to .58). Thus, when participants hated the chain-branded focal brand and the same-subcategory option had a better quality reputation, the money-back guarantee provided the necessary assurance that the brand with the better quality reputation would be the safer option despite being from the same subcategory as the hated brand (M higher quality with guarantee = 2.87, SD  = 1.98, M higher quality without guarantee = 4.30, SD = 1.32, t(104) = 4.45, p < .001, d = .91).       121 A18. Telecommunication Scenario (Study 7) Imagine you were recently in Canada for about eighteen months for business. During your stay, you subscribed to a local cell phone service company called Fido. You signed up with Fido after doing some research on Canadian cell phone companies because Fido offers plans tailored for customers that frequently need to call the US.  While drinking your morning coffee and reading the news, you stumble across a newspaper article about the Canadian telecommunications industry and its major players.       Please click --> to go to the article.      Article: Canadian telecommunications industry - Page 1      The Canadian telecommunications industry has changed radically in the past 10 years, as data-hungry customers with smart devices consume ever more bandwidth. Over this period, operators have expanded their service portfolios and overhauled their price plans to meet explosive demand, while simultaneously upgrading their network capabilities. A recent consumer report suggests that operators’ offerings vary significantly in terms of plans, extras, network coverage, and service. However, customer satisfaction is very similar across providers. Despite all the changes over the years, the main industry structure has remained stable as the market is divided into two categories: established companies (such as Fido and Telus) and newer players (such as Wind). While the established players such as Fido and Telus have more local branches, the newer companies have a leaner structure.        122 Article: Canadian telecommunications industry - Page 2 (featuring the manipulation of salience of the competitive rivalry)  Low salience condition High salience condition Together the three companies – Fido, Telus, and Wind – are the market leaders and account for more than 70% of the market. While there are high levels of competition in this industry, the most direct competitors in the industry are Fido and Telus, which are both established companies.   Together the three companies – Fido, Telus, and Wind – are the market leaders and account for more than 70% of the market. While there are high levels of competition in this industry, the two key rivals in the industry are the two established operators, Fido and Telus, which fiercely battle for customers.     This rivalry has become increasingly heated as the CEO of Fido has repeatedly emphasized that the key strategic priority of Fido is taking customers from Telus.     Reading this article reminded you of your own experience with FIDO. Their bills were significantly above what you had been told they would be. There were a bunch of hidden fees that they never explained when you signed up and you felt they had exploited you. You were charged a $2 fee for having the service, and then another 25 cents per US call. You are disgusted with FIDO and feel like they were underhanded in selling their service. When you called the customer service department to complain, the sales representative seemed     123 completely uninterested in helping you. He told you that you signed the agreement and if you are upset then you should have read it more carefully before you signed. His tone was very rude and it was obvious he didn’t care at all. You felt full of contempt for him. In that moment, you realized that FIDO is an evil company and that you truly hate them.  The next time you come to Canada for business you are determined to change providers.    Dependent variable   Imagine that you back in Canada. In order to carry your number over to your new provider, you have to let FIDO know which provider you picked.  Please put yourself in this situation and indicate which of the two phone providers you would pick.  [presentation order was counterbalanced] - Telus  - Fido        124 A19. Hatred Versus Anger Follow-Up Study  Below is a full description of the study comparing the experience of anger (without hatred) and hatred referenced in the future research section. Method Participants and design. Four hundred participants (51% female, Mage = 37.4), recruited from MTurk, were asked to list brands toward which they felt hatred or anger. At the beginning of the study, I highlighted the definitions drawn from the Merriam Webster dictionary for hatred (www.merriam-webster.com/dictionary/hate) and anger (www.merriam-webster.com/dictionary/ anger). Specifically, I used the following definition for hate: “intense hostility and aversion usually deriving from a sense of injury; extreme dislike or disgust”. Whereas anger was defined as: “a strong feeling of displeasure and usually of antagonism”. After reading these definitions, I prompted participants to think of brands they felt hatred and/or anger for. The questions were presented on the same page and participants could either indicate, “Yes” and write down the brand name or indicate that they did not feel hate and/or anger for any brand. On the subsequent page, I invited participants to briefly describe their feelings and experiences with the brand toward which they felt hatred and/or anger. Subsequently, I measured the dependent measures for participants that indicated that they felt hatred and/or anger for a brand (N = 297) with two questions for each listed brand asking them how likely they would be to a) “purchase from a close competitor of the brand as a way of getting back at BRAND NAME” and b) “purchase from a company that was quite different from BRAND NAME as a way of protecting myself from being wronged again” on seven point scales (1 = “not at all likely”. 7 = “extremely likely”). After this task, participants provided demographic information.     125 Results  Preferences for competing brands. To analyze how feeling hatred versus anger influences consumer motivations and preferences for competing brands, I first analyzed the preferences among consumers who indicated that they felt hatred or anger for a brand. Those who indicated that they hated a brand (N = 172, 43% of the sample) were significantly more likely to choose a more distant competitor in order to protect themselves (M = 5.63, SD = 1.72) than to choose a close competitor in order to get back at the focal brand (M = 4.60, SD = 2.31, t(171) = 5.23, p < .001, d = .51). In contrast, among those who reported feeling anger for a brand (N = 260, 65% of the sample), I did not find a significant differences between the motivation to choose a more distant competitor in order to protect themselves (M = 4.61, SD = 2.12) and the motivation to choose a close competitor as a way of getting back at the focal brand (M = 4.56, SD = 2.03, t < 1, p > .73, d = .03). To further probe this pattern, I also conducted a subsample analysis using only those consumers who indicated that they felt both anger and hate toward different brands (N= 135, 34% of the sample). In this subsample, I found when consumers hated a brand they were significantly more likely to choose a distant competitor (M = 5.67, SD = 1.71) than when they experienced anger toward a brand (M = 4.56, SD = 2.03, t(134) = 5.44, p < .001, d =.59).  Discussion  My follow-up study directly contrasted hatred with anger and used the actual feelings consumers harbor toward different brands to show that hatred, but not anger, leads to a greater tendency to protect oneself by choosing distant competitors of the focal brand. The propensity of hatred to lead consumers to prioritize self-protection via choosing distant competitors emerged both between-subjects (i.e., by examining all consumers who experienced hatred or anger) as well as within-subjects (i.e., by focusing on those who experienced both feelings.) 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0378066/manifest

Comment

Related Items