i BRAND MIND PERCEPTION & MORAL JUDGMENTS OF BRAND BEHAVIOR: HOW PERCEIVED LEADERSHIP INFLUENCES CONSUMER ATTITUDINAL RESPONSES TO A BRAND’S WRONGDOING by HU XIE A DISSERTATION Presented to the Department of Marketing and the Graduate School of the University of Oregon in partial fulfillment of the requirements for the degree of Doctor of Philosophy June 2017 ii DISSERTATION APPROVAL PAGE Student: Hu Xie Title: Brand Mind Perception and Moral Judgments of Brand Behavior: How Perceived Leadership Influences Consumer Attitudinal Responses to a Brand’s Wrongdoing This dissertation has been accepted and approved in partial fulfillment of the requirements for the Doctor of Philosophy degree in the Department of Marketing by: T. Bettina Cornwell Chairperson Conor Henderson Core Member Lynn Kahle Core Member Gerard Saucier Institutional Representative and Scott L. Pratt Dean of the Graduate School Original approval signatures are on file with the University of Oregon Graduate School. Degree awarded June 2017 iii © 2017 Hu Xie iv DISSERTATION ABSTRACT Hu Xie Doctor of Philosophy Department of Marketing June 2017 Title: Brand Mind Perception and Moral Judgments of Brand Behavior: How Perceived Leadership Influences Consumer Attitudinal Responses to a Brand’s Wrongdoing How we communicate about brands and companies has changed. CEOs have come into the spotlight of brand communications but little marketing research offers holistic knowledge about CEOs as brand endorsers. This research investigates how CEO endorsers influence consumer attitudes toward a brand differently from conventional endorsers (e.g., celebrities and athletes). Further, this research examines underlying mechanisms that determine consumer responses to CEOs as brand endorsers and especially consumer moral judgments of a brand’s wrongdoing. Building on research on brand endorsers and brand equity, as well as drawing theoretical support from research on leadership, anthropomorphism and mind perception, this dissertation proposes a moderated mediation model of CEO endorser effects on consumer moral judgments. Brand endorsers for decades have been viewed as essentially communicating via three characteristics: attractiveness, expertise and trustworthiness. This dissertation identifies perceived leadership as an additional endorser dimension elicited from a CEO-brand endorser. Further, this dissertation introduces brand mind perception into marketing research and finds that perceived leadership positively influences consumers’ perception of brand mind, which in turn determines consumers’ v moral judgments. Boundary conditions are explored and include endorser-brand relationship and crisis controllability. Two sets of studies provide empirical support. The first set defines and develops the scale of perceived leadership including item generation (Study 1), item purification (Study 2), and scale confirmation (Study 3). The second set tests the hypotheses in the conceptual model. Two exploratory studies first find preliminary evidence of that perceived leadership differs from existing endorser dimensions by its effects on moral judgments (Study 4), and that mind perception is possible for a brand and can be enhanced by CEO association (Study 5). Study 6 shows positive effects of CEO endorsers on consumer attitudes by communicating perceived brand leadership. Study 7 investigates a brand-wrongdoing scenario and shows that perceived brand leadership yields negative results for a brand by increasing blame and reducing forgiveness; Study 8 demonstrates these relationships are mediated by brand mind perception. Study 9 shows that the inspiring aspects of perceived leadership can enhance perceptions of brand mind (to feel and experience), thus reducing consumers’ blame. Theoretical and managerial implications are discussed. vi CURRICULUM VITAE NAME OF AUTHOR: Hu Xie GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED: University of Oregon, Eugene University of Wisconsin, Madison Shanghai Jiao Tong University, Shanghai DEGREES AWARDED: Doctor of Philosophy, Marketing, 2017, University of Oregon Master of Science, Marketing, 2013, University of Oregon Master of Business Administration, Marketing, 2009, University of Wisconsin Bachelor of Arts, Arts Administration, 1998, Shanghai Jiao Tong University AREAS OF SPECIAL INTEREST: Social Perception Branding and Brand Relationship New Technology Moral Judgment Consumer Welfare and Prosocial Consumption PROFESSIONAL EXPERIENCE: Graduate Teaching and Research Fellow, Department of Marketing, University of Oregon, 2011 - 2017 Marketing Manager, Johnson & Johnson, 2009 - 2010 Category Marketing Manager, Coca-Cola, 2005 - 2007 Brand Activation Manager, Coca-Cola, 2003 - 2004 Assistant Brand Manager, Nestle, 1998 - 2002 GRANTS, AWARDS, AND HONORS: Beta Gamma Sigma, University of Oregon Chapter, 2017 vii University Dissertation Fellow, University of Oregon, Lundquist College of Business, 2016 Robin and Roger Best Research Award, University of Oregon, Lundquist College of Business, 2016 AMA-Sheth Doctoral Consortium Fellow, London Business School, 2015 Robin and Roger Best Teaching Award, University of Oregon, Lundquist College of Business, 2015 Robin and Roger Best Research Award, University of Oregon, Lundquist College of Business, 2015 Merle King Smith Marketing Scholar Award, University of Oregon, Lundquist College of Business, 2011 to present MBA Merit Scholarship with Full Tuition Remission, University of Wisconsin – Madison, Wisconsin School of Business, 2007-2009 PUBLICATIONS: Xie, H., Minton, E. A., & Kahle, L. R. (2016). Cake or fruit? Influencing healthy food choice through the interaction of automatic and instructed mental simulation. Marketing Letters, 27(4), 627-644. Xie, H., & Cornwell, T. B. (2016). Why voters don’t seem to forgive Clinton, while Trump gets a free pass. The Conversation. Retrieved from https://theconversation.com/why-voters-dont-seem-to-forgive-clinton-while-trump-gets- a-free-pass-68020 viii ACKNOWLEDGMENTS I would like to express the deepest appreciation to my committee chair Dr. Bettina Cornwell. Without her encouragement and coaching, this dissertation would not have been possible. Her support and persistence have helped me through the doctoral training to become a scholar. I would also like to thank my committee members, Dr. Conor Henderson, Dr. Lynn Kahle, and Dr. Gerard Saucier for their support and generosity in sharing their knowledge and experience with me. Discussions and conversations with them have always been inspiring and encouraging. I am profoundly grateful for the tremendous love and support I receive from my family. They have always been there for me, giving me greatest happiness and strength. ix For my Mom whose blessings have made me smile every day ☺ x TABLE OF CONTENTS Chapter Page I. INTRODUCTION .................................................................................................. 1 An Emerging Phenomenon of CEOs as Brand Endorsers ..................................... 1 Dissertation Research Overview and the Conceptual Model ................................ 2 General Contribution ............................................................................................. 6 Studies Overview and Key Findings...................................................................... 7 II. DEVELOPMENT OF THE CONCEPTUAL MODEL AND HYPOTHESES...... 10 Brand Endorsers and Endorser Dimensions .......................................................... 10 CEOs as Brand Endorsers and Existing Leadership Research .............................. 11 Perceived Leadership and General Consumer Attitudes ....................................... 18 Perceived Leadership and Brand Mind Perception ................................................ 20 Mind Perception and Anthropomorphism ....................................................... 20 Research Gap in Brand Anthropomorphism .................................................... 22 Brand Mind Perception and Perceived Leadership from a CEO endorser ...... 23 Brand Mind Perception and Individual Differences ........................................ 25 Brand Mind Perception and Consumer Moral Judgment ...................................... 27 Boundary Conditions ............................................................................................. 29 III. EMPIRICAL DESIGN, ANALYSES, AND FINDINGS ..................................... 32 Perceived Leadership Scale Development ............................................................. 32 Study 1- Item Generation and Selection ......................................................... 32 Study 2 - Exploratory Factor Analysis ............................................................ 33 Study 3 - Scale Confirmation .......................................................................... 37 xi Chapter Page Exploratory Studies ................................................................................................ 44 Study 4 – Exploring Perceived Leadership ..................................................... 45 Study 5 – Exploring Brand Mind Perception .................................................. 50 Investigating the Proposed Moderated Mediation Model ...................................... 59 Study 6 - Perceived Leadership and Brand Attitudes ...................................... 59 Study 7 - Perceived Leadership and Consumer Moral Judgment .................... 65 Study 8 - Brand Mind Perception as the Link to Moral Judgment .................. 72 Study 9 - Considering Inspiring & Dominant Dimensions of Leadership....... 80 IV. CONCLUSION AND GENERAL DISCUSSION ................................................ 95 Research Contribution ........................................................................................... 97 Research Limitations and Future Research Opportunities..................................... 100 Managerial Implications ........................................................................................ 103 APPENDICES ............................................................................................................. 105 A. Study 4 Stimuli.................................................................................................. 105 B. Study 6 Stimuli .................................................................................................. 106 C. Study 7 Stimuli – News Article......................................................................... 107 D. Study 8 Stimuli ................................................................................................. 108 E. Study 8 Stimuli - News Article ......................................................................... 109 F. Study 9 Stimuli .................................................................................................. 110 G. Study 9 Stimuli - News Article ......................................................................... 111 H. Measures Summary ........................................................................................... 112 REFERENCES CITED ................................................................................................ 117 xii LIST OF FIGURES Figure Page 1. Dissertation Research Scope, Focus, and Theoretical Support .............................. 3 2. The Moderated Mediation Model of CEO Endorsers’ Effects on Consumer Moral Judgment ................................................................................................................ 5 3. Study 6: Simple Mediation Paths ........................................................................... 63 4. Study 6: Alternative Mediation Paths .................................................................... 64 5. Study 7: Effects on Moral Judgments of a Brand’s Wrongdoing .......................... 69 6. Study 7: Moderated Mediation Paths ..................................................................... 71 7. Study 8: Interactive Effect on Perception of Brand Mind ..................................... 78 8. Study 8: Moderated Mediation Paths ..................................................................... 79 9. Study 9: Perceived Leadership Scores ................................................................... 85 10. Study 9: Perceptions of Brand Agency and Experience Minds ............................. 88 11. Study 9: Moderated Mediation Paths ..................................................................... 90 12. Study 9: Effects of Perceived Leaderships on Consumer Moral Judgment........... 92 xiii LIST OF TABLES Table Page 1. A – Study 2: Item Correlation Matrix B – Study 2: EFA Matrix Output ........................................................................... 36 2. Study 3: CFA Model Fits ....................................................................................... 39 3. Study 3: Construct Validity Test Results ............................................................... 42 4. Study 4: Construct Validity Test Results ............................................................... 47 5. Study 4: Moral Judgments of Political Figures ..................................................... 49 6. Study 5: Results of GLM Model of Perception of a Brand as Having a Mind ...... 55 7. Study 5: Results of Logistic Model of Perception of a Brand as Having a Mind ....................................................................................................................... 57 8. Study 7: Table of Means of Moral Judgments ....................................................... 68 9. Study 7: Results of Moderated Mediation Paths by Mediators ............................. 72 10. Study 9: Table of Means of Endorser Perceptions................................................. 84 11. Study 9: Results of Moderated Mediation Path Analysis ...................................... 91 12. Study 9: Results of Regression Model of Perceptions of Brand Mind on Moral Judgments .............................................................................................................. 94 1 CHAPTER I INTRODUCTION An Emerging Phenomenon of CEOs as Brand Endorsers CEOs are among the most powerful people associated with branding. Some CEOs are recognized as highly associated with their brands such as Steve Jobs and Tim Cook with Apple, Bill Gates with Microsoft, and Charles Schwab with Charles Schwab Financial Services, while others remain in the background. Research in management, economics, and corporate finance has thoroughly examined the impact on business performance of CEO-related factors (e.g., Haleblian & Finkelstein, 1993; Parrino, 1997; Yermack, 1995). Surprisingly, little marketing research has addressed the potential effects of CEOs on brand perceptions and consumer attitudes, despite a long history of CEOs working as brand endorsers (e.g., Paul Newman for Newman’s Own, John Schnatter for Papa John’s Pizza, and Charles Schwab for Charles Schwab Financial Services). One of a few exceptions is the research of O’Connell and O’Sullivan (2011), however, they investigated how customer satisfaction will impact CEO compensation rather than CEOs’ effects on brands. The need for marketing research on CEO effects has become apparent as more and more CEOs have come into the spotlight as the faces of their brands in marketing communications. These CEOs extend their roles from managing the internal and external workings of the organization to acting as a spokesperson for their brands and products, and thus communicating with consumers directly in marketing communications, not simply through corporate actions. For example, the marketing communication strategy of 2 Tesla is rare and exceptional for brands. Advertising Age reported “Tesla Motors has no advertising, no ad agency, no CMO, no dealer network. And that’s no problem” (McCarthy, 2013). Tesla earns media and advertising exposure by Elon Musk doing interviews with media and customers to communicate about projects and products (McCarthy, 2013). As a similar example in a different industry, in 2016, Doug McMillon became the first Walmart CEO to star in a TV commercial (Sozzi, 2016). Many questions have arisen from this change in the role of industry leaders. Do CEO endorsers influence consumer attitudes toward a brand in the same way as conventional brand endorsers such as celebrities and athletes? If not, what are the underlying mechanisms determining these relationships? What are the positive and potential negative consequences on consumer attitudes? This dissertation aims to respond to these questions. Both theoretical academic research and industry practice should benefit from more understanding of the phenomenon where CEOs as leaders perform the function of brand endorsers. Dissertation Research Overview and the Conceptual Model This section summarizes the main proposals and findings as well as their theoretical support of this dissertation work. Examples of prior research and the numbers of chapters and sections for detailed discussion in the dissertation body are indicated in parentheses. First, this research (Chapter II, section 1) reviews literature in marketing which shows that brand endorsers benefit brands by transferring an endorser’s associations to a brand and thus contributing to brand equity building, resulting in positive consumer 3 attitudes toward a brand (Keller, 1993; Park, Jaworski, & MacInnis, 1986). This finding provides a research foundation in marketing literature for this dissertation work to build on. Figure 1 shows the foundation, as well as the research focus and main proposals of this dissertation work (i.e. the top part of diagram with black blocks) as well as related theoretical supports (i.e. the theories mentioned with arrows). Figure 1. Three endorsers’ dimensions such as trustworthiness, expertise, and attractiveness have been found to have critical effects on consumer perceptions and thus the persuasiveness of marketing communications (Ohanian, 1990). However, this dissertation postulates that these dimensions may not be able to capture the unique characteristics of CEOs who are distinctive leaders. This research identifies that perceived leadership is the missing characteristic found in endorsers now commonly engaged in brand and product promotion. After extensive research in leadership literature (Chapter II, section 2), this dissertation finds existent leadership research focuses on individual personality traits and behaviors of leaders in influencing subordinates and organizational performance, and 4 these leadership qualities are either self-reported by a leader or evaluated by their subordinates based on their direct experience with a leader (e.g., Bass & Avolio, 1994; Wayne, Shore & Liden, 1997). In contrast, little research examines leadership perceived by consumers and its consequences on consumer attitudes. Thus, this dissertation proposes perceived leadership is defined as the leadership qualities perceived but not necessarily experienced by the observer; who can be external parties such as consumers, customers or the general public, outside of an organization. This research develops a perceived leadership scale to measure perceived leadership (Chapter III, section 1). Besides general consumer attitudes toward a brand as a consequence of communications by a leader (Chapter II, section 3), this dissertation is interested in consumer moral judgments as one focal outcome. Theoretically, communication by a leader, just like other communications about a brand, may influence many associations held by consumers (Keller 1993). These associations in turn are expected to influence the ways in which consumers respond to brands. Based on research on mind perception in psychology (e.g., Gray & Wegner, 2009), this dissertation introduces mind perception to marketing research, and identifies perception of brand mind as a critical link between perceived leadership of CEO endorsers and consumer moral judgments (Chapter II, sections 4 and 5.). Based on the work of Gray and Wegner (2009) and in application to the marketing context, this research defines perception of brand mind as the extent to which consumers perceive a brand as having mental capabilities, and includes the concepts agency mind as the ability to think, plan, and act toward a goal, and experience mind as the ability to have feelings and sensory experiences. Findings in anthropomorphism (e.g., Epley, Waytz, & Cacioppo, 2007) and brand association (e.g., 5 Keller 1993) provide support to the proposal that the perceived leadership induced by a CEO endorser enables the transfer of a leader mentality to perception of the brand (Chapter II, section 4). Brand mind perception is expected to play a decisive role in shaping consumer moral judgment toward a brand (Chapter II, section 5). Mind perception research shows that when people attribute minds to others, people identify another’s capabilities to take action (Waytz, Gray, Epley, & Wegner, 2010). Thus, this research predicts that higher perception of brand mind in general suggests greater responsibilities and blameworthiness of a brand for wrongdoing. In sum, this dissertation proposes a moderated mediation model of CEO endorsers’ effects on consumer moral judgment: that perceived leadership initiated by CEOs as brand endorsers positively influences consumer perceptions of brand mind, which in turn determines consumers’ moral judgments of a brand’s wrongdoing. This framework incorporates potential moderators and boundary conditions including endorser-brand relationship and crisis controllability (Chapter II, section 5). See Figure 2 for model graphic. This visual outline aims to guide further reading of this dissertation. Figure 2 6 General Contribution This dissertation extends brand endorser research by providing a new criterion, perceived leadership, which captures the unique characteristics of CEOs as brand endorsers and thus differentiates them from conventional endorsers. In addition, perceived leadership provides a new perspective (i.e., consumers’ perceptions and consequences) into leadership research that traditionally focuses on effective leadership in managing subordinates and enhancing organizational performance. This research contributes to applied research by introducing the concept of brand mind perception and examining the implications that stem from it. This dissertation proposes that consumers can perceive a brand’s mind, and such perception of a brand’s mind is distinct from the attribution to brands of other humanlike characteristics previously investigated in brand anthropomorphism research. Its importance lies in evaluation of a brand’s social capabilities and judgments of that brand’s actions. Investigating these topics will enhance understanding of how consumers may imbue brands with human-like mental capabilities that may eventually alter consumers’ moral judgments of a brand. Studies Overview and Key Findings To test the proposed conceptual model, a series of studies has been conducted to show how perceived leadership of a CEO as a brand endorser can influence perception of brand mind, and how brand-mind perception consequently predicts consumers’ moral judgments toward a brand. 7 Two sets of studies are incorporated to provide empirical evidence supporting the proposals of this research. Perception of brand mind is an extension of mind perception research that has measures of mind perception available for adaption; however, there is no existing scale that is appropriate to measure perceived leadership. Therefore, the first set of studies defines the perceived leadership construct and develops the perceived leadership scale. This first set of three studies addresses perceived leadership scale item generation (Study 1), item purification (Study 2), and factorial structure confirmation (Study 3). These studies test the construct validity of the proposed perceived leadership scale by exploring its relationships with relevant traditional endorser characteristics such as trustworthiness, expertise, and attractiveness. As an overview, the second set of studies (Studies 4 – 9) examines the relationships among perceived leadership from CEO endorsers, brand mind perception, and consumer attitudes and moral judgments as consequences. These studies test hypotheses in the conceptual model, demonstrating perception of brand mind plays a critical role in bridging perceived leadership effects and shaping consumer moral judgments. The results not only show the importance of including the perceived leadership scale in endorser research, but also confirm the superior predictive power of brand-mind perception on consumer moral judgments of brand behaviors. Furthermore, this set of studies explores how utilizing a CEO as a brand endorser will differ from featuring conventional endorsers such as athletes and employees, and provide useful implications to marketing practitioners. This second set of studies (4-9) serves several purposes. Before testing the hypotheses in the conceptual model, this research presents two exploratory studies that 8 investigate foundational questions about perceived leadership (Study 4) and brand mind perception (Study 5). Specifically, Study 4 examines how people would respond when political figures are involved in wrongdoings, providing preliminary evidence of the positive relationship between perceived leadership and people’s blame and responsibility attribution. The concept of brand mind perception is adapted from mind perception research, thus new to marketing research. To examine this adaptability as well as gain more understanding of brand mind perception, Study 5 explores whether consumers can perceive a brand as having a mind and, if so, what individual differences, as well as brand factors may influence people’s tendency to perceive a brand’s mind. Study 5 provides early evidence that CEO association may influence perception of brand mind. Study 6 explores positive effects of using a CEO (vs. athlete) as a brand endorser in marketing communications, showing that the enhanced perception of leadership leads to higher consumer attitudes toward a brand and its advertisements. Study 7 investigates a fictitious scenario where a brand is required to recall a product due to toxic materials. The results confirm that perceived leadership stemmed from featuring a CEO as a brand endorser can lead to more blame and less forgiveness for a brand from consumers. In addition, the study reveals a boundary condition where the effects of perceived leadership on moral judgments are attenuated when a wrongdoing situation can be controlled, thus indicating a high attribution of responsibility (Russell, 1982) and resulting in a ceiling effect. Study 8 investigates the link between perceived leadership and moral judgments and finds that perception of brand mind mediates this relationship. In addition, Study 8 shows that whether a CEO (or an employee) is related to a brand (i.e., a brand’s own 9 CEO rather than a CEO from another brand) moderates whether brand-mind perception can be altered by perceived leadership. Based on source credibility theory (e.g., Hovland et al., 1953), this research postulates that separating a CEO from a brand disqualifies the conjecture of a CEO’s perceived leadership to a brand’s mind perception, thus eliminating downstream moral judgments on a brand. Finally, Study 9 explores how manipulating the two dimensions of perceived leadership—either inspiring-leadership or dominant-leadership—can influence perceptions of brand mind differently. Indeed, the results show that consumers perceive higher experience in the mind of a brand with inspiring leadership rather than dominant leadership, and consequently, consumers are more likely to forgive and less likely to blame the former than the latter. This result provides further understanding of how the sub-dimensions of brand-mind perception such as experience and agency impact moral judgments in different ways. This dissertation contributes to marketing research by enhancing understanding of how bringing a brand’s CEO into the spotlight of a brand’s marketing communications may alter consumers’ perceptions and attitudes toward a brand, especially moral judgments about brand wrongdoing. In addition, the findings and implications can be used to guide marketing practitioners in brand-building as well as in brand crisis situations. The final section (Chapter V) of this dissertation recognizes research limitations and provides suggestions for future research directions. 10 CHAPTER II DEVELOPMENT OF THE CONCEPTUAL MODEL AND HYPOTHESES Brand Endorsers and Endorser Dimensions The goal of building a brand is to establish a brand position in the market and clearly differentiate a brand from its competitors, thus enhancing its market performance (Shocker & Srinivasan, 1979; Oxenfeldt & Swan, 1964). This differential effect relies on the strength of brand equity that, from the consumer perspective, is undergirded by important brand associations (Keller, 1993). Brand associations can be created by consumers’ direct experience with a product or service (e.g., product attributes and benefits, packaging design, and pricing) or inferred from links with other entities (e.g., sponsored events, country of origin, spokesperson or endorser) (Cornwell, Roy, & Steinard, 2001; Keller, 1993; Park, Jaworski, & MacInnis, 1986). Brand spokespersons and endorsers have been found to influence consumer perceptions, brand image and brand equity (Spry, Pappu, & Cornwell, 2011). When a brand is identified with another entity such as a brand endorser, consumers infer that the brand shares associations and characteristics with that entity (Keller, 1993). For example, brand endorsers may enable a transfer of global associations such as attitude or credibility to the brand and its offerings (Keller, 1993; Ohanian, 1990). Different groups of endorsers such as consumers, celebrities, or well-known athletes have been examined for their influence on consumer brand decisions (e.g., Erdogan, 1999; Carlson & Donavan, 2008). 11 Three endorsers’ dimensions have been found to have critical effects on consumer perceptions and thus the persuasiveness of marketing communications. These three dimensions are trustworthiness, expertise, and attractiveness (Ohanian, 1990). Trustworthiness (i.e., the degree of confidence in the communicator's intent to communicate the assertions) and expertise (i.e., the extent to which a communicator is perceived to be a source of valid assertions) are the factors influencing the perceived credibility of a communicator, and thus underlie the idea of source credibility (Hovland, Janis, & Kelley, 1953); attractiveness is proposed based on the source attractiveness model (McGuire, 1985), which acknowledges the power of likeability, familiarity and similarity in determining consumer attitudes (Ohanian, 1990). In marketing communication and advertising literature, trustworthiness, expertise, and attractiveness are well established and seemingly “the” dimensions of endorsers. CEOs as Brand Endorsers and Existing Leadership Research In the wake of CEOs coming into the spotlight of brand communications, this research anticipates the existence of unique characteristics of CEOs as brand endorsers, and suggests that these unique characteristics may not be captured by the established endorser dimensions. “A CEO is the leader of a firm, responsible for its overall operations and performance” (businessdictionary.com). The distinctive responsibilities and characteristics of being a leader of the firm being endorsed should distinguish a CEO from conventional endorsers such as celebrities and athletes, as well as employees and customers. Thus, this dissertation proposes that perceived leadership is one missing brand endorser characteristic. 12 Leadership is a unique property of extraordinary individuals (Carlyle, 1849). Research in social psychology as well as research in management and organizational behavior has shown the importance of leadership in influencing subordinates’ (e.g., employees’) performance and shaping an organization’s behavior (e.g., Bass & Avolio, 1994; Wayne, Shore & Liden, 1997). However, little research has focused on leadership from the consumer perspective in terms of how a CEO’s leadership is perceived by consumers and thus influences consumer attitudes toward a brand. To differentiate the leadership examined by prior research from the leadership that is examined in this dissertation, the former is classified as “actual” leadership and the latter as “perceived” leadership. The “actual” leadership research focuses on leader behaviors that are effective in motivating subordinates and enhancing organizational performance (e.g., Bass & Avolio, 1994). For example, research finds that effective leaders who can inspire followers to work towards common goals are the ones who offer followers, something beyond just the feeling of working for self-interest. Effective leaders provide inspiring vision and motivation, as well as trust and encouragement to take actions; this leadership is coined transformational leadership (Burns, 1978; Bass 1990). In addition, an effective leader is found to give clear assignments and make him-or-her-self available to support followers (Stogdill, 1963). The “actual” leadership research, in addition, is interested in what personality traits can predict a leader’s effectiveness. For example, established personality scales such as Big Five mini markers (Saucier, 1994), have been leveraged by researchers to examine characteristics that may transform into leader qualities and thus differentiate 13 leaders from others (Quirk & Fandt, 2000). Bono and Judge (2004) find “extraversion” as the strongest and most consistent correlate of transformational leadership among the Big Five personality traits. Another example from research in corporate finance shows that risk-taking CEOs are highly associated with risk-taking firms; thus, showing how a CEO’s personal attributes can shape corporate policies and organizational behaviors (Cain & McKeon, 2016). A variety of leadership scales have been proposed, tested, and applied in previous research to examine “actual” leadership. These existing scales concern actual leadership traits and characteristics that influence the behaviors of subordinates and organizations. Thus, the existing scales of leadership involve either self-reported evaluations by managers or other people in leadership positions, or evaluations of them by their subordinates. Therefore, these scales are more appropriate in contexts where actual dyadic relationships of supervisors and subordinates exist rather than contexts where third parties’ perceptions (e.g., consumers) are of interest, as in this research. Specifically, in the context of marketing communications where CEO endorsers demonstrate leadership qualities, existing scales are not useful or applicable. Thus, this dissertation proposes a perceived leadership as a new scale to capture how the public (e.g., consumers at large) perceives a leader’s (e.g., a CEO’s) qualities. The following discussion reviews major leadership research as well as related scales in the literature and demonstrates the need for perceived leadership research with a new scale, especially in consumer research. Transformational Leadership and Multifactor Leadership Questionnaire (MLQ). Research on leader behavior identifies two prominent leadership styles, transactional 14 leadership and transformational leadership (Bass, 1985). Transactional leadership is built on a series of exchanges or bargains for values, while transformational leadership implies the alignment of values, goals, and beliefs (MacKenzie, Podsakoff, & Rich 2001; Morhart, Herzog, & Tomczak, 2009). Leadership is powerful in influencing employee performance and shaping organizational behavior. Both transactional and transformational leadership have been found to motivate frontline employees in eliciting brand-building behavior (Morhart, Herzog, & Tomczak, 2009). The transformational and transactional leadership scale (i.e., MLQ) developed by Bass (1985) and Avolio, Bass and Jung (1990) is a prominent scale used especially in management and organization behavior research. The re-examined and re-defined scale (Avolio, Bass, & Jung 1990) includes a total of 36 items loaded by five higher-level leadership dimensions. Specifically, the scale asks employees to evaluate their direct- report supervisors on a frequency scale (anchored on 1 = not at all, to 4 = frequently, if not always) in terms of charisma (12 items), intellectual stimulation (4 items), individualized consideration (4 items), contingent reward (4 items), active management (4 items), and passive avoidance (8 items). Employees evaluate their supervisors on these items in terms of frequency. For example, the items for charisma include “___ goes beyond self-interest,” “___ talks enthusiastically,” and “___ arouses awareness;” and the items for intellectual stimulation include “___ suggests different angles,” and “___ seeks different views.” These items would not be applicable to the context of investigating consumer and public perceptions. For example, when encountering a print advertisement by Walmart featuring its CEO, Doug McMillon, consumers would not know whether he “teaches and coaches” (an item for individual consideration), “rewards employee’s 15 achievement” (an item for contingent reward), or “tracks your mistakes” (an item for active management). The MLQ scale has been adapted for marketing research by Morhart, Herzog, and Tomczak (2009) to examine how brand-specific leadership of managers promotes employees’ brand building behaviors. Specifically, they asked employees to rate their direct report managers on a 30-item brand-specific leadership scale. For example, the contingent reward dimension has 4 items, including “Points out what I will receive if I do what is required from a brand representative,” and the charisma dimension has 4 items, including “Talks optimistically about the future of our corporate brand.” The research by Morhart and colleagues (2009) is so far the only research, found by this dissertation work, that has tackled leadership-related questions in marketing, but still focuses on the dyadic relationship between managers and subordinates. The Ohio State Leadership and the Leader Behavior Description Questionnaire (LBDQ). This scale is also called the Ohio State Leadership Scale, due to its development by a group of interdisciplinary researchers including psychologists, sociologists, and economists at Ohio State University. The series of Ohio State leadership studies have made an important contribution to the literature on leadership by identifying two types of leadership behaviors facilitating goal accomplishment: initiating structure and consideration (Schriesheim & Stogdill, 1975). Initiating structure summarizes leader behaviors that are task-oriented and positively influence group members to work, cooperate and achieve goals as a group; consideration includes people-oriented behaviors that promote support and care to group members (Schriesheim & Stogdill, 1975). 16 The related LBDQ scale has been applied in hundreds of research studies. The book Leader Behavior: Its Description and Measurement, edited by Stogdill and Coons (1957) describes the details of scale development and testing with the original 150 items. A simplified version of the LBDQ has been tested and analyzed using the factor structural method, resulting in two dimensions as initiating structure and consideration with a total of 20 items to measure leader behavior (Schriesheim & Stogdill, 1975). Again, these leader behavior indicators would in most instances be un-available to general consumers outside an organization. The Leadership in Sports Management and Leadership Scale for Sports (LSS). The LSS scale was developed by Chelladurai and Saleh (1980) in the field of sports management, to capture coaching behaviors. Leadership theory is highly relevant in sports because sports teams are basically organizations (Ball, 1975) that are concerned with the development of people. LSS contains a total of 40 items mapping on five factors: training and instruction, democratic behavior, autocratic behavior, social support, and positive feedback. Athletes respond to these items by rating their coaches. Some example items are “lets his athletes share in decision making” (an item for democratic behavior) and “helps the athletes with their personal problems” (an item for social support). Opinion Leadership and Its Scale (OLS). Social scientists, including consumer and marketing researchers have long been interested in opinion leaders. In consumer research, opinion leadership refers to consumers’ behaviors in which they influence others’ purchase decisions (Mowen, 1990). A variety of scales have been developed to measure opinion leadership, including the Myers and Robertson’s (1972) scale, the Rogers’ (Rogers & Cartano, 1962) scale, and the King and Summers’ (1970) scale. 17 The King and Summers scale is frequently applied in consumer and marketing research, and its original version contains a set of seven self-report items. Childers (1986) revised and refined the King and Summers scale, and the revised scale shows higher internal consistency and reliability, as well as better construct validity (Childers, 1986). Flynn, Goldsmith, and Eastman (1994) used factorial analysis to test its dimensionality and further reduce the scale to six items. For example, the scale asks people to respond to questions such as: “In general, do you talk to your friends and neighbors about ___? (1=never, 5=very often)” or “When you talk to your friends and neighbors about ___, do you: (1=give very little information, 5=give a great deal of information” (Childers, 1986). The opinion leadership and its scale should not be confused with the perceived leadership proposed by this research. Perceived leadership focuses on evaluating traits that consumers (or the public) typically perceive that leaders possess. By contrast, opinion leadership scales are self-designating assessments of a person’s behavioral tendency in sharing opinions and thus influencing others in a particular, usually product- related context. Leader Personality Traits and Attributes Framework (LTA). Personality traits are important in predicting effective leadership, and established personality scales such as Big Five mini markers (Saucier, 1994) have been used by researchers to examine leadership qualities (Quirk & Fandt, 2000). Building on personality research, Zaccaro, Kemp, and Bader (2004) created a framework to guide personal traits and attributes research in effective leadership. Research shows that personality variables and attributes are effective in the prediction of leader effectiveness (e.g., Judge, Bono, Ilies, & Gerhardt, 2002; Peterson, Smith, Martorana, & Owens, 2003; see Zaccaro, Kemp, & 18 Bader, 2004, for a review). Moreover, contexts are important in determining what traits are linked to leadership effectiveness. For example, the need for prudent, bold, flexible leadership becomes secondary when resources are abundant and there is less competition (Judge, Piccola, & Kosalka, 2009). Perceived Leadership and Effects on Brand Perception & Consumer Attitudes In contrast to the “actual” leadership that is measured based on the perspectives of internal parties such as managers themselves or subordinates and employees, perceived leadership is defined as the leadership qualities perceived but not necessarily experienced by the observer; who can be external parties such as consumers, customers or the general public, outside of an organization. Accordingly, the outcome variables of perceived leadership are related to consumer perceptions and attitudinal responses, rather than effectiveness as measured by employee or organizational performance. In addition, after an extensive review of the previous research, it appears that no existing leadership scales are appropriate and applicable to measure the perceived leadership proposed by this dissertation. Therefore, this research will first develop a new scale, the perceived leadership scale. Studies 1 – 3 in Chapter III will present the process and results of perceived leadership scale development. In a context of marketing communication such as advertising, this research predicts that featuring a CEO as a brand endorser will lead to positive consumer attitudes toward a brand. Consumers can infer a brand’s associations and characteristics from its links with other entities such as brand endorsers (Keller, 1993; Ohanian, 1990). Cognition research provides additional support by offering a possible cognitive process 19 of priming. A prime activates associated knowledge (Collins & Loftus, 1975) in a way similar to the availability heuristic, which assesses top-of-mind awareness and thus influences perceptions and decisions (Yi, 1990). For example, when participants are primed with high- (or low-) class, participants evaluate fictitious car names as high- (or low-) class (Herr, 1989). As noted, leadership is a unique and admirable property of extraordinary individuals (Carlyle, 1849). Because of the strong associations of CEOs with leadership, it is expected that making a CEO a brand endorser salient in marketing communications will activate thoughts related to leadership and consequently influence consumer perceptions of a brand and their attitudes toward that brand. Therefore, this research proposes (see Figure 2 for hypotheses mapping in the conceptual model): H1: A CEO as a brand endorser is perceived by consumers as having higher leadership qualities than conventional brand endorsers. H2: Featuring a CEO as a brand endorser in marketing communications will positively influence consumer perception of a brand as a leading brand in its industry, and perceived leadership mediates this relationship. H3: Featuring a CEO as a brand endorser in marketing communications will positively influence consumer attitudes toward a brand, and perceived leadership mediates this relationship. 20 Perceived Leadership and Brand Mind Perception This dissertation further proposes that the perceived leadership induced by a CEO endorser may also enable the transfer of a leader mentality to perception of the brand. That is, this research predicts that perceived leadership elicited from a CEO as a brand endorser will enhance the perception of a brand’s mind. However, before this Chapter discusses more details of the prediction, it starts with a review of research on mind perception, from which this dissertation adopts the concept of mind perceptions, and research on anthropomorphism, which enables the adaptation of mind perceptions to a brand. Mind Perception and Anthropomorphism Perceiving another’s mind is critical in social life (Waytz et al., 2010). One indication of children maturing is that they develop “theory of mind” through social activities to differentiate others’ minds from their own and predict what is in the other’s mind (Baron-Cohen, Leslie, & Frith, 1985). Two streams of research have developed that are closely linked because both concern how people interpret the mental processes and behavior of others, including nonhuman others: one is mind perception theory and the other is anthropomorphism. Mind perception research is unique for its focus on differentiating human from nonhuman subjects. Most research investigates the mind as a single dimension that works as an underlying continuum and that accommodates an entity between “no mind” and “full mind” (e.g., Morewedge, Preston, & Wegner, 2007). However, recent mind 21 perception theory has argued that people intuitively perceive other minds in terms of two dimensions: agency and experience (Gray et al., 2007; Waytz et al., 2010). Agency refers to the mental capacity to plan and act (e.g., thinking, memory, taking actions); experience refers to the mental capacity to feel and sense (e.g., feeling hunger or pain, experiencing joy) (Gray et al., 2007). These two dimensions map out and compare mind perceptions of different human and nonhuman entities. For example, the mind of an adult human is perceived as high in both dimensions; however, the mind of an animal such as a chimpanzee, dog, or frog is perceived as high in experience but low in agency (Gray et al., 2007). It implies that the one-dimensional perception of mind (i.e., from no mind to full mind) may essentially emphasize the agency mind of the two-dimensional model. In sum, mind perception research investigates how people perceive other minds and how these minds vary (Gray et al., 2007; Waytz et al., 2010). Anthropomorphism research examines people’s tendencies to imbue nonhuman objects with humanlike characteristics, motivations, intentions, and emotions (Epley et al., 2007; Premack & Woodruff, 1978). A tree can be viewed as wise, a broken computer as being mad, and a car as an old friend. Anthropomorphizing a machine may attribute human characteristics to it so that people may perceive it as having thoughts and intentions (Kim & McGill, 2011; Waytz et al., 2010). Thus, anthropomorphism in fact encourages people to perceive a nonhuman entity as having a human mind (Waytz et al., 2010), and essentially influences mind perception of a nonhuman subject. Anthropomorphism research provides opportunities to study changes in mind perception and their consequences to existing mind perception research. 22 Research Gap in Brand Anthropomorphism Marketing research shows that consumers attribute humanlike characteristics to products and brands as well (Aggarwal & McGill, 2007). Products and their advertising designed with humanlike features, such as a human body-shaped bottle or a car’s front headlamps looking like eyes, may trigger people to see a product as more human (Aggarwal & McGill, 2007). Research shows that anthropomorphism can have positive effects by enabling a sense of efficacy with nonhuman entities or increasing emotional bonding with them (Kim & McGill, 2011). When a product has human-like features that match activated positive human schemas, people will show more favorable attitudes toward the product (Aggarwal & McGill, 2007) as well as experience more pleasure and arousal (Landwehr, McGill, & Herrmann, 2011). Anthropomorphizing a product may encourage consumers to apply social expectations and beliefs that they might not otherwise assign to the product (Kim & McGill, 2011). For example, when anthropomorphizing and treating a car as if it were alive, people give less consideration to its actual quality or current state and are more reluctant to replace it (Chandler & Schwarz, 2010). Most anthropomorphism research in marketing focuses on the product and is interested in whether a product “looks” or “appears” human. Marketing research has yet to examine brand mind perceptions, the attribution of human-like mental capabilities to a brand or the consequences of anthropomorphizing brands in the form of attributing minds to them. This absence is in spite of the fact that mind perceptions are critical to social interactions (Waytz et al., 2010), and that consumers have been shown to treat brands as social partners (Fournier & Alvarez, 2012). 23 Prior research suggesting and demonstrating, but not explicitly examining, brand mind perceptions has shown that brands can be perceived to have personalities (Aaker, 1997), that consumers form social relationships with their brands (Fournier, 1998), and that consumers develop commitment and loyalty to their brands (Oliver, 1999). These findings imply that consumers can perceive a brand as humanlike, allowing for deeper social connections. This stream of research on brand social relationships builds from the assumption that brands can be treated as people (Fournier & Alvarez, 2012) and thus regularly directs research attention to downstream consequences. The findings showing consumer-brand social interactions provide evidence of consumer perceptions of a brand’s social capabilities, supporting the proposal of this dissertation that consumers may perceive a brand as having a mind. Brand Mind Perception and Perceived Leadership from a CEO Endorser This dissertation adopts the concepts of mind perception in psychology and introduces them to marketing research. Building on mind perception research, this research defines brand mind perception as the extent to which consumers perceive a brand as having mental capabilities. In keeping with past basic research (e.g., Morewedge et al., 2007), this research first examines brand mind perception as an overall continuum (from no mind to full mind). Secondly, as in past research of mind perception (Gray et al., 2007), brand mind perception is explored as having two dimensions: agency (e.g., to think, plan, and act toward a goal) and experience (e.g., to feel, sense, and experience). In contrast to previous research on anthropomorphism, which examines 24 perceptions in terms of a generic consideration of a product appearing or being more or less humanlike, this research offers a more nuanced perspective that anchors on humanlike mental capabilities and examines the variances of a brand’s mind as either a global measure or as having two sub-dimensions: agency and experience. What can influence perception of brand mind? This research predicts that in a context of marketing communication, featuring a CEO as a brand endorser will influence the perception of brand mind. A person’s identity is largely related to mind (Strohminger & Nichols, 2014). Mentality and intellectual abilities have been recognized as important characteristics of leaders by both the literature (e.g., House & Aditya, 1997; Bass, 1990; Kirkpatrick & Locke, 1991) and industry (e.g., Zook & Allen, 2016). When a leader such as a brand’s CEO comes to the spotlight of a brand’s marketing communication, the strong association between a leader and the brand enables the transfer of a leader mentality to perception of mind. Thus, in the context of marketing communications with a CEO as a brand endorser, this research predicts that perceived leadership of a CEO endorser will positively influence perception of a brand’s mental capabilities. H4: Featuring a CEO as a brand endorser in marketing communications will positively influence perception of brand mind, and perceived leadership mediates this relationship. 25 Brand Mind Perception and Individual Differences Is it commonplace to perceive a brand as having a mind? If so, what are the characteristics of those individuals most likely to perceive the mind of a brand? Because brand mind perception examines the phenomenon in which consumers attribute a mind to nonhuman subjects, therefore, it is expected that the individual differences that motivate the anthropomorphism tendency will also be important in terms of brand mind perception. This section discusses three individual differences that have been found in anthropomorphism research. Thus, this section summaries these as proposals of direct application in a new context, rather than new hypotheses. To explain when people are likely to anthropomorphize and when they are not, Epley, Waytz, and Caccioppo (2007) find that a person’s desire for social contact and affiliation plays a critical role. They predict that people are more likely to anthropomorphize when they are in higher need of social connection. For example, lonely people are more likely to talk to a pet as they would to a human companion. Accordingly, this research proposes that social needs will predict the likelihood of seeing a brand’s mind as well, and specifically, when people are lonely and lack social connection, they are more likely to see a brand as having a mind. Social needs can relate to an individual’s self-construal that explains self-related behaviors, and there are two types of construal: independent self-construal and interdependent self-construal (Escalas & Bettman, 2005). Independent self-construal indicates a tendency to emphasize the personal self in terms of unique personal traits but deprecate others; in contrast, interdependent self-construal emphasizes the social self and relationships with others (Brewer & Gardner, 1996). For example, research has shown 26 that Westerners (e.g., Europeans and North Americans) tend to have more independent self-construal, while Easterners (e.g., Asians) tend to have more interdependent self- construal (Markus & Kitayama, 1991). However, these two self-construal levels coexist, and it is possible to prime consumers on one of them (Brewer & Gardner, 1996). This research predicts that when a person thinks of the social self, a person has higher social needs and thus is more likely to perceive a brand as having a mind. Anthropomorphism provides people with a sense of understanding other subjects or environments, especially in confusing or uncertain situations (Epley et al., 2007). For example, when a computer crashes, a person with little technology knowledge tends to perceive the computer as being mad and losing its mind like a human, because such a belief provides a sense of understanding about what is happening to the computer. However, people who have high need for cognition are inclined towards effortful cognitive activities and desire to understand situations in a reasonable way (Cohen, Stotland, & Wolfe, 1955). They may reject a shortcut solution that may include anthropomorphizing and seek logical explanations. This seesaw reflection is captured when a person says, “Wait a moment; a brand doesn’t have a mind because it is not human!” Therefore, people with high need-for-cognition will be less likely to see a brand as having a mind. Together, three individual differences are summarized as: P1: Loneliness has a positive effect on the tendency to perceive a brand’s mind. P2: Interdependent self-construal has a positive effect on the tendency to perceive a brand’s mind. 27 P3: Need-for-cognition has a negative effect on the tendency to perceive a brand’s mind. Brand Mind Perception and Consumer Moral Judgment Perceiving another’s minds is crucial in forming moral judgments (Gray & Wegner, 2009). Mind perception provides an intuitive explanation of others’ actions by revealing desire, intentions, and consciousness (Epley, Schroeder, & Waytz, 2013). When people attribute minds to others, people identify another’s capabilities to take action, including wrongdoing (Waytz et al., 2010). Thus, higher mental capabilities in general suggest greater responsibilities and blameworthiness for wrongdoing. Further research in mind dimensions provides additional support by showing agency mind as a driving force in the relationship between mind perception and moral judgment. Specifically, research finds that different mental capabilities correlate with different moral judgments; blame for wrongdoing correlates more with agency (ability to think, plan, and make decisions) than experience (mental ability to feel and sense) (Gray, Gray, and Wegner 2007). Agency mind emphasizes a person having a capacity to act in a goal-oriented direction and thus remain blameworthy, whereas experience mind is associated with a person seeming more vulnerable as the target of wrongdoing (Gray & Wegner, 2009) and less capable of doing, therefore less blameworthy (Gray & Wegner, 2011). 28 Building on these findings, this research predicts that brand mind perception will influence consumer moral judgments towards a brand’s actions in a similar way. It is expected that enhancing brand mind perception will lead to stringent moral judgments of a brand. For example, when perceiving a brand as having a mind, consumers will judge a brand as more responsible for wrongdoing, thus attributing more blame to and being less forgiving of a brand. However, emphasizing or enhancing experience mind may lead moral judgment in an opposite direction, toward less blaming and more forgiving. Thus, H5: Enhancing brand mind perception will increase consumers’ attribution of blame to a brand for its wrongdoing (i.e., brand mind perception has a positive relationship with consumers’ likelihood of attributing blame to a brand for wrongdoing). H6: Enhancing perception of brand experience mind will reduce consumers’ attribution of blame to a brand for its wrongdoing (i.e., brand experience mind has a negative relationship with consumers’ likelihood of attributing blame to a brand’s wrongdoing). In sum, the brand mind perception and consumer moral judgment model in this dissertation proposes that when a CEO enters the spotlight of marketing communications as a brand endorser, the perceived leadership induced by the prominence of a CEO will enhance consumer perception of brand mind. However, in a brand wrongdoing situation, enhanced brand mind perception will lead to more stringent moral judgments from 29 consumers, meaning there will be more blame and less forgiveness of a brand. In contrast to a brand advertising situation where featuring a CEO as a brand endorser positively influences consumer perception of a brand and attitudes toward a brand, a CEO brand endorser may lead to undesirable consequences in a brand wrongdoing situation. Boundary Conditions Are these predictions regarding the effects of perceived leadership and brand mind perception on consumer moral judgments always applicable? This dissertation acknowledges two boundary conditions: one is the relationship between a CEO and a brand, and the other is the controllability of processes and causes involved in wrongdoing (i.e., crisis controllability). Leaders may be sought after as endorsers for many brands other than their own. First, this research argues that not all perceived leadership from a CEO endorser would enhance perception of a brand’s mind and influence moral judgments of a brand. This relationship depends on whether a CEO is related to a brand or not (i.e., a brand’s own CEO or a CEO from a different brand), and a CEO from a different brand may not produce the same results in influencing consumer moral judgments. This thinking is supported by source credibility theory (Hovland et al., 1953) and research in responsibility attribution (Russell, 1982). Specifically, consumers may still evaluate the perceived leadership of both CEOs at a similar level in spite of their relationship with a brand; nonetheless, a CEO is a CEO and he or she is a leader. However, when a CEO is a brand’s own CEO, the close connection between the two entities facilitates the 30 association inference between them, whereas a loose connection may block the reference due to lack of source credibility (Keller 1993; Hovland et al., 1953). In addition, the disconnection suggests less involvement of a CEO in a brand’s decision making, thus less mentality transfer. Thus, this research predicts that a CEO’s relationship with a company moderates the effects of a CEO as a brand endorser on brand mind perception. Specifically, H7: When an endorser is a CEO but without a relationship to the brand, the effects of perceived leadership from a CEO on brand mind perception will be attenuated, as will the downstream effect on moral judgments of a brand’s wrongdoing. The second boundary condition is related to how people perceive the controllability of a crisis. Research has identified that people make attributions by assessing whether the cause of a crisis is controllable; when a crisis is considered as more controllable, a higher degree of responsibility will be assigned (McAuley, Duncan, & Russell, 1992; Russell, 1982). For example, people consider a crisis with an internal origin (e.g., material contamination at a brand’s own facility) as more controllable than a crisis with an external origin (e.g., at a supplier’s facility), and thus perceive the former situation as one in which the brand is more responsible and blameworthy (Coombs & Holladay, 2011; Liu, Austin, & Jin, 2011). Therefore, this research predicts a ceiling effect for brand endorsers’ influence on consumer moral judgments: when crisis 31 controllability is high, consumers attribute full responsibility to a brand, so that the effects from brand endorsers on brand moral judgments disappear. H8: When a crisis is perceived as highly controllable, the effects of perceived leadership induced by a CEO endorser on consumer moral judgments will disappear. The next Chapter will discuss a total of nine studies in detail to provide empirical evidence to support the proposals and hypotheses. These studies can be grouped as two sets of studies. The first set of studies (Studies 1 – 3) defines and develops the construct of perceived leadership scale because no existing scale appropriately measures perceived leadership. The second set of studies (Studies 4 – 9) examine the relationships among perceived leadership from CEO endorsers, brand mind perception, and consumer attitudes and moral judgments as consequences. 32 CHAPTER III EMPIRICAL DESIGN, ANALYSES AND FINDINGS Studies 1-3: Perceived Leadership Scale Development Perceived leadership induced by a CEO endorser is identified as a powerful brand factor in shaping perception of brand mind. However, the literature has little discussion on perceived leadership, thus requiring construct development and validation. This section includes three studies to develop a perceived leadership scale, which is important in the proposed conceptual model and thus allows model testing in the next section. Study 1 - Item generation and selection An initial pool of items was generated through an online survey and a review of previous research (Churchill, 1979). The online survey was utilized to explore how people perceive the characteristics of leaders. Participants (n = 61, Mage = 32.9, 41% female) from the Amazon mTurk were randomly assigned to think about in general either (1) a leader, (2) a person with leadership qualities, or (3) a CEO, and they were asked to list adjectives that would describe such a person. These three variations were used to elicit a broad range of characteristics. Next, participants were asked to write down the names of people whom they think could represent leaders, or display leadership qualities. The analysis of trait words began by examining their frequency of mention across the three conditions. During this process, words with similar meanings were grouped together and represented by the most frequently mentioned word. For example, 33 “intelligent” and “smart” were grouped together, and “intelligent” was chosen to represent the group because it was mentioned more frequently than “smart.” Similarly, “courageous” was chosen for “courageous” and “brave,” and “capable” for “capable” and “competent.” The words that were neither adjective nor indicating personal traits (e.g., Boss, Command, Money) were eliminated. Furthermore, the list generated from the literature review (seven items) was compared and contrasted with the list from the survey, and the duplicated ones were removed, resulting in eight additional items. Based on the guidance of the leaders’ personal traits and attributes framework by Zaccaro and colleagues (2004), sematic differential items, which indicate stable personal traits, were examined and selected under an additional consideration of their applicability in situations where evaluators have no direct experiences with a leader. The whole process generated a total of 15, seven-point sematic differential items. Political figures (e.g., Barack Obama, Ghandi, Martin Luther King, Jr.), religious leaders (e.g., Jesus Christ, Moses, the Pope) and business executives (e.g., Steve Jobs, Bill Gates, Elon Musk) were among most mentioned names as a leader or a person who displays leadership. These results were expected since people in a top position or role of influence, such as leader of a country or a movement, religious figures, and corporate CEOs, are commonly considered as leaders. Study 2 - Exploratory Factor Analysis The objective of Study 2 was to examine empirically the possible underlying factors in the fifteen leadership semantic differentials. Participants (n = 153, Mage = 34.27, 48.4% female) from the Amazon mTurk online panel participated in the study for 34 a small cash incentive. The study asked participants how they evaluate brand endorsers. Specifically, participants were randomly assigned to think about: (1) a celebrity, (2) an athlete, or (3) a CEO in general, and rate this endorser on the 15 items of personal traits from Study 1. They then responded whether they were thinking about a specific person (yes or no), and if yes, they were asked to provide the name of this person. Finally, participants were asked to indicate how much they perceived this endorser as a leader. Specifically, the question asked, “Please indicate to what degree you agree or disagree with the statements below (1 = strongly disagree; 7 = strongly agree): 1. Celebrities in general are leaders; 2. Celebrities in general typify what leaders are.” “Celebrities” were replaced by “athletes” or “CEOs” depending on the conditions they were assigned. Method and Results. The first step was to examine the factorability of the data set as well as the 15 PL items. The sample size of 153 participants, using listwise deletion and providing10 cases per variable was acceptable for the factor analysis minimum amount requirement (Field 2013). Further, all 15 items were at least correlated at .3 with one other item, suggesting reasonable factorability (see Table 1 - Panel A). The Kaiser- Meyer-Olkin measure of sampling adequacy was .87, and thus, above the commonly recommended value of .6, and Bartlett’s test of sphericity was significant (χ2 (105) = 1100.18, p < .00). These criteria for the factorability all suggest that factor analysis was applicable. Finally, except for the item “masculine,” the communalities were all above .3, further confirming that each item shared certain common variance with other items. Second, exploratory factor analysis (EFA) was conducted with principle axis factoring extraction and promax oblique rotation. An oblique rotation was used because it was expected that the hypothesized dimensions describing the structure would be inter- 35 correlated. The initial results of Eigen values (> 1) suggested three underlying dimensions. However, the pattern of factor loadings suggested that only two of three extracted dimensions were uniquely defined, and the third dimensions had insufficient primary loadings and was difficult to interpret. Visual inspection of the scree plot also confirmed that a two-factor solution was appropriate, explaining over 53.55% of the variance in the items. Thus, the two-dimension solution was retained. Further examining EFA results at the item level resulted in removal of the items with primary loading less than .600, and items that cross-loaded onto the other factor. This process reduced the scale to a total of 7 items (see Table 1 – Panel B). The first factor was labeled as “inspiring” and includes four items: inspiring, responsible, effective, and courageous. All these items in this factor reflect the motivating and influential traits that consumers perceive a leader has. The second factor was labeled as “dominant” and includes three items: dominant, authoritative, and powerful, indicating the status and power-related traits that consumers perceive a leader has. Two factors were correlated (r = .643, p < .05) with satisfactory internal consistencies for both factors (inspiring: α = .81; dominant: α = .74; total α = .83) within the scale development requirements proposed by Nunnally (1978). 36 Table 1 Table 1. A – Study 2: Item Correlation Matrix Table 1. B – Study 2: EFA Matrix Output In addition, how participants would consider three types of endorsers in terms of the belief of being a typical leader (r = .88) was investigated. A one-way ANOVA revealed a significant difference among endorsers (F(2,150) = 14.47, p < .00), and Tukey post hoc test showed that CEOs (M = 5.50) were perceived as leaders more than celebrities (M = 4.09, p <.00) and athletes (M = 4.83, p <.05). The seven items from EFA analysis were aggregated as the perceived leadership scale, and a one-way ANOVA was conducted with the scale as a dependent variable. Again, the results revealed a significant difference among endorsers (F(2,150) = 9.61, p < Responsible Masculine Intelligent Effective Transformat ional Inspiring Influential Dedicated Courageous Confident Authoritative Capable Dominant Decisive Powerful Responsible 1.000 .092 .642 .520 .467 .453 .249 .463 .456 .332 .351 .554 .279 .471 .341 Masculine .092 1.000 .061 .186 .039 .115 .178 .351 .158 .184 .139 .178 .280 .255 .202 Intelligent .642 .061 1.000 .493 .481 .431 .449 .365 .393 .344 .527 .557 .425 .598 .395 Effective .520 .186 .493 1.000 .474 .510 .518 .599 .492 .462 .360 .704 .490 .537 .422 Transformati onal .467 .039 .481 .474 1.000 .375 .373 .445 .427 .405 .234 .394 .206 .278 .171 Inspiring .453 .115 .431 .510 .375 1.000 .350 .451 .573 .199 .287 .483 .198 .304 .306 Influential .249 .178 .449 .518 .373 .350 1.000 .374 .244 .466 .389 .467 .554 .406 .457 Dedicated .463 .351 .365 .599 .445 .451 .374 1.000 .408 .470 .221 .570 .341 .443 .348 Courageous .456 .158 .393 .492 .427 .573 .244 .408 1.000 .261 .304 .466 .307 .426 .372 Confident .332 .184 .344 .462 .405 .199 .466 .470 .261 1.000 .422 .422 .436 .424 .506 Authoritative .351 .139 .527 .360 .234 .287 .389 .221 .304 .422 1.000 .288 .427 .449 .549 Capable .554 .178 .557 .704 .394 .483 .467 .570 .466 .422 .288 1.000 .381 .437 .385 Dominant .279 .280 .425 .490 .206 .198 .554 .341 .307 .436 .427 .381 1.000 .596 .510 Decisive .471 .255 .598 .537 .278 .304 .406 .443 .426 .424 .449 .437 .596 1.000 .436 Powerful .341 .202 .395 .422 .171 .306 .457 .348 .372 .506 .549 .385 .510 .436 1.000 Correlation Matrix 1 2 3 1 2 3 1 2 3 Effective .795 -.124 .174 Effective .765 .366 .277 Responsible .772 .082 -.171 Intelligent .746 -.074 -.453 Capable .749 .517 .506 Inspiring .720 -.110 .030 Capable .739 -.203 .107 Responsible .749 .512 .206 Effective .653 -.057 .008 Decisive .697 .181 -.072 Intelligent .748 .737 .102 Capable .639 .024 .340 Dedicated .682 -.182 .391 Inspiring .663 .365 .309 Courageous .617 .003 .056 Responsible .674 -.294 -.212 Courageous .644 .421 .333 Intelligent .585 .496 -.351 Dominant .635 .450 .065 Transformational .620 .580 .505 Transformational .560 .097 .317 Influential .627 .222 .076 Dominant .403 .719 .518 Dedicated .530 -.140 .510 Powerful .619 .339 -.025 Authoritative .430 .703 .192 Authoritative -.006 .742 -.092 Confident .607 .214 .124 Decisive .575 .696 .406 Dominant -.214 .728 .332 Courageous .604 -.230 -.004 Powerful .438 .683 .421 Powerful -.069 .650 .200 Inspiring .586 -.323 .000 Influential .483 .612 .483 Decisive .177 .536 .120 Authoritative .575 .305 -.280 Confident .465 .578 .507 Influential .060 .466 .276 Transformational .556 -.278 -.027 Dedicated .668 .397 .693 Confident .051 .421 .322 Masculine .269 .154 .287 Masculine .165 .216 .412 Masculine -.092 .118 .408 Structure Matrix Factor Factor Matrix Pattern Matrix Factor Factor 37 .00). Tukey post hoc test showed that CEOs (M = 5.75) scored significantly higher on the perceived leadership scale than celebrities (M = 5.06, p <.00) and athletes (M = 5.27, p <.01). The correlation analysis between perception of being a typical leader and the proposed leadership scale showed a significant positive correlation (Pearson Correlation r = .638, p < .00). This result is expected and reflects the idea that if a person is thought of as a leader, she or he should have leadership traits as indicated by the perceived leadership scale. Study 3 - Scale Confirmation This study aims to use a structural model method to confirm the scale produced in the exploratory scale study (Study 2). College students (N = 135) participated in this study for exchange of credit. They were randomly assigned to evaluate (1) a CEO, (2) a celebrity, or (3) an athlete, each from a pool of five candidates in each category. The participants first were shown the name, title and a picture of the assigned candidate, and then they responded to the seven items of semantic differential perceived leadership scale that were obtained from the prior EFA study. If a person is considered as a leader, this person should be perceived to have leadership qualities. Based on this thinking, perceived leadership scale should have concurrent validity with perception of a person as a leader. Therefore, participants were directly asked about their general perceptions of the person as a leader, and the objective was to help assess concurrent validity of the proposed perceived leadership scale. Specifically, they were asked to what extent they agreed with the following statements about the person they evaluated: 1) the person is a leader; 2) this person typifies what a 38 leader is; 3) this person has leadership qualities (anchored on 1 = strongly disagree; 7 = strongly agree). The study included additional questions about the original three endorser dimension measures taken from Ohanian (1990), including trustworthiness (5 items: dependable, honest, reliable, sincere, trustworthy), expertise (5 items: expert, experienced, knowledgeable, qualified, skilled), and attractiveness (5 items: attractive, classy, beautiful, elegant, sexy). Method and Results. The proposed leadership scale, with two dimensions loading on seven items, was tested by performing the confirmatory factor analysis with AMOS 23 (Arbuckle, 2013). The results indicated a good model fit (χ^2 =16.189, df = 13, p = .239) with all model fit indices exceeding recommendations (CFI = .995, NFI = .974, GFI = .966, RMR = .041, RMSEA = .043, PCLOSE=.523). To test whether the proposed structure is better than alternatives, the next analysis was to compare the model fits of the two-dimension model (proposed) with two other confirmatory models including a zero-factor null model (model A) and a single-factor model where all seven items were loaded from one factor (model B) (See Table 2). The model A showed a poor model fit (χ2 = 623.001, df = 21, p = .000), and model B showed acceptable model fit (χ2 = 27.327, df = 14, p = .017; CFI = .978, NFI = .956, GFI = .941, RMR = .057, RMSEA = .084, PClose = .110). The Chi-Square difference test showed that the difference between the proposed model and Model B was marginally significant (Δ χ2 = 11.138, df = 1, p = .08); and the difference between the proposed model and the null model was significant (p < .00). In general, the model fit indices suggested that the proposed two-dimension model fit better than other models. - 39 - Table 2 Study 3: CFA Model Fit - 40 - Further, AVE and CR for both dimensions in the proposed model were calculated to explore their convergent validity and composite reliability. AVE for “inspiring” was .620, and .708 for “dominant,” exceeding the criterion of .50; and CR for “inspiring” was .867 and .879 for “dominant,” exceeding the criterion of .70 (Fornell & Larcker, 1981). Using SPSS, the overall inter-item reliability was accessed, showing high level of consistency among seven items (α = .918). It was expected to establish concurrent validity between the proposed perceived leadership and general belief as a leader (α = .939). This expectation was based on the notion that a leader in general would demonstrate the perceived leadership traits, while a person with these traits would be generally esteemed as a leader. A bivariate correlation analysis showed that the two constructs were strongly positively associated (r = .699, p < .001). Thus, the perceived leadership scale exhibits concurrent validity in the leader perception domain. Convergent (with Expertise and Trustworthiness) and Discriminant (with Attractiveness) Validity. The established scale of measuring trustworthiness (5 items: dependable, honest, reliable, sincere, trustworthy), expertise (5 items: expert, experienced, knowledgeable, qualified, skilled), and attractiveness (5 items: attractive, classy, beautiful, elegant, sexy) (Ohanian, 1990) has been well applied in accessing endorsers’ perceived characteristics. In the same domain, the proposed perceived leadership scale examines leadership traits and characteristics of a person or endorser. A positive relationship is expected between perceived leadership and trustworthiness and between perceived leadership and expertise. As demonstrated by the significant concurrent validity between perceived leadership and general belief as a - 41 - leader, a person with higher perceived leadership traits will more likely to be considered as a leader. A leader is the one who leads and guides others (Merriam-Webster), and a leader is the one on whom others can depend. Thus, a leader in general should own a high level of trust from others, and usually a leader earns leadership by certain expertise and skills. The convergent validity tests with CEO data confirmed this expectation by showing that overall perceived leadership scale (α = .918) was positively related to trustworthiness (α = .934; β = .407, p < .01) and expertise (α = .948; β = .502, p < .01). Similar results were replicated by using celebrity data and athlete data. See Table 3 for detail. - 42 - Table 3 Study 3: Construct Validity Test Results CONSTRUCT VALIDITY TESTS FOR PERCEIVED LEADERSHIP SCALE Type of assessment PL PL - Inspiring PL - Dominant PL PL - Inspiring PL - Dominant PL PL - Inspiring PL - Dominant Concurrent validity Belief as a leader .699*** .645*** .688*** .618*** .587*** .545*** .626*** .605*** .605*** Convergent / Discriminant validity: Trustworthy .407*** .484*** .245*** .288*** .321*** .067 .336*** .401*** .215** Expertise .502*** .400*** .585*** .567*** .481*** .574*** .461*** .433*** .439*** Attractiveness .060 .088 .013 .052 .071 .018 .172*** .117 .218** Note: PL = Perceived Leadership ** Statistically significant at p < .05 *** Statistically significant at p < .01 CEO Celebrity Athlete - 43 - Examining the sub-dimensions of perceived leadership (PL), PL-inspiring dimension is expected to have a higher positive relationship with trustworthiness than PL-dominant dimension. This proposal acknowledges the assumption that genuine trust is more likely to be gained or related to inspiring qualities of a person such as being courageous and effective rather than dominant qualities such as position or power. A person who has power is not necessarily trusted by others; respectively, a person who is trusted might not have the authority to make final decisions. The analysis of CEO data showed that PL – inspiring (α = .866) had a higher positive relationship with trustworthiness (β = .484, p < .01) than PL – dominant (α = .879; β = .245, p < .01). The same analysis procedure with celebrity data and athlete data respectively except PL – dominant was not related with trustworthiness for celebrity data (α = .771; β = .067). Further, it was expected that both dimensions—inspiring and dominant—are positively related with expertise at a similar level. Expertise indicates the advanced skill or knowledge a person has comparing to peers. Thus, an expert can be inspiring and may hold authority and power as an opinion leader especially in a relevant field. Reciprocally, being an inspiring leader or a powerful leader may suggest that they own certain skills and knowledge. The data supported this proposition. For all three data sets, PL – inspiring was positively related to expertise (CEO data: α = .948; β = .400, p < .01; Celebrity data: α = .925; β = .481, p < .01; Athlete data: α = .903; β = .433, p < .01), as was PL – dominant to expertise (CEO data: β = .585, p < .01; Celebrity data: β = .574, p < .01; Athlete data: β = .439, p < .01). In contrast with the propositions on trustworthy and expertise, PL is not expected to directly relate to attractiveness scale. This prediction is based on careful examining the - 44 - items of attractiveness including “attractive, classy, beautiful, elegant, sexy,” which focus more on appearances and physical appeals. Although attractiveness enhances persuasion (McGure, 1958), these traits should not be confused with the charisma of leadership that taps on the behavioral attributes such as “talking optimistically” and “expressing confidence” (Avolio, Bass, & Jung, 1999; Rubin, Munz, & Bommer, 2005). Both CEO and celebrity data confirmed this proposal by showing no significant relationships between PL overall, PL – inspiring, PL – dominant and attractiveness (all ps > .05). However, when participants evaluated athletes, leadership was related to attractiveness (β = .172, p < .01), and this can be explained by the nature of being an athlete who competes to be the best in terms of physical power and fitness. Studies 4 – 5: Exploratory Studies Before being able to test the hypotheses in the conceptual model regarding the relationships among perceived leadership, brand mind perception and consumer moral judgments, two exploratory studies were conducted to investigate foundational questions. To further test the construct validity as well as to provide preliminary evidence of the potential relationship between perceived leadership and people’s blame and responsibility attributions, Study 4 examines how people would respond when political figures are involved in wrongdoing. Study 5 explores the extent to which consumers can perceive a brand as having a mind and, what individual differences, as well as brand factors may influence people’s tendency to perceive a brand’s mind. - 45 - Study 4 – Initial Exploring Perceived Leadership and Moral Judgments The purpose of this study is to explore the relationship between perceived leadership and moral judgments, as well as to provide further evidence confirming nomological network of the proposed construct when compared with exiting endorser measures. Design and Method. Two hundred and nine college students participated in this study for course related credit. This study was conducted in early 2016 during the U.S. presidential campaign when the major political parties (Democratic and Republican) had just narrowed to two candidates per party. Participants were asked to evaluate randomly one of five U.S. political figures including the current leader, President Barack Obama, the Democratic candidates for the presidential nomination Hillary Clinton and Bernie Sanders, as well as the Republican candidates for the presidential nomination Donald Trump and Ted Cruz. These political figures were presented to study participants with their names, titles, and photos. See Appendix A for details. The visual presentations were selected for their consistency in dressing (formal), facial expression (smiling), and facing direction (toward right side of viewers) to minimize potential effects from these visual presentation aspects on attitudes. Participants first rated their assigned politicians on the 7-item perceived leadership scale, then on traditional endorser measures including trustworthy (5 items), attractiveness (5 items), and expertise (5 items). As in Study 3, participants rated the person assigned to them in terms of general belief as a typical leader (3 items). See Appendix H for measure summary. - 46 - To explore their moral judgments, participants were asked about their moral emotion response (disappointment, 1 item), moral behavior tendency (blame and forgiveness, 1 item each), and moral intention judgment (intention, 1 item). Specifically, participants were asked to indicate to what extent they would agree or disagree with the statements: 1) “If this person did something wrong, I would be very disappointed” 2) “I would blame this person if he or she did something wrong” 3) “I would forgive this person if he or she apologized for a wrongdoing” (1 = strongly disagree; 7 = strongly agree). In addition, they responded to the question “If this person did something wrong, how likely would you believe he or she did it intentionally?” by a 7-point Likert scale anchored on 1 = extremely unlikely; 7 = extremely likely. Finally, participants indicated their general attitudes toward the person they rated by three items (unfavorable/favorable, bad/good, dislike/ like; (Mitchell & Olson, 1981)) on a seven-point bipolar scale. Results. First, the relationships between perceived leadership and the general belief as a leader as well as three traditional endorser measures were analyzed to examine construct validity of perceived leadership as in previous studies. The analyses yielded the same results as Study 3. See Table 4 for details. - 47 - Table 4 Study 4: Construct Validity Test Results Specifically, the results offer convergent evidence of the ability of the leadership scale to capture leadership characteristics. Specifically, significant correlation coefficients were found between perceived leadership (PL) and the general belief as a leader (α = .920, β = .807, p < .01), as well as between the sub-dimensions of perceived leadership with the general belief as a leader (PL – inspiring: β = .838, p < .01; PL – dominant: β = .580, p < .01), supporting the proposed construct as highly relevant in the leadership perception domain. Further, the analyses of the relationships with traditional endorser perceptions yielded the similar results as Study 4. Specifically, the data confirmed the convergent validity between PL and trustworthiness (α = .969, β = .340, p < .01) and expertise (α = .969, β = .502, p < .01) as well as discriminant validity between PL and attractiveness (α = .881; β = .052, n.s.). As expected, the sub-dimension PL – - 48 - inspiring had a positive relationship with trustworthiness (β = .528, p < .01) and expertise (β = .381, p < .01) but not attractiveness (β = .057, n.s.), while PL – dominant was only positively related with expertise (β = .586, p < .01). This study focuses on the potential effects of perceived leadership on consumer moral judgments in general rather than on the comparative attitudes regarding political candidates. Thus, three models of regression analyses were conducted with collapsed data of all candidates, and used disappointment, blame, forgiveness, and intention as an outcome variable respectively. Model 1 used general attitudes (α = .986) as a sole predictor, Model 2 added traditional endorser attributes (trustworthiness, attractiveness, and expertise) as well as the perceived leadership as new predictors. Model 3 used the sub-dimensions (PL - inspiring and PL - dominant) to replace the perceived leadership to detect whether differences between two sub-dimensions existed. See Table 5 for the result summary. - 49 - Table 5 Study 4: Moral Judgments of Political Figures Predicting Results: Regression Model of Perceived Leadership On Moral Judgements Blame ForgiveDisappointment If this person did something wrong, I would be very disappointed. Intention If this person did something wrong, how likely would you believe he or she did it intentionally? I would blame this person if he or she did something wrong. I would forgive this person if he or she apologized for a wrongdoing. General Attitudes General Attitudes .33 *** .08 .03 -.20 *** -.35 ** -.30 ** .65 *** .32 *** .31 *** -.56 *** -.28 ** -.24 ** Existing Endorser Attributes Trustworthy .07 -.05 -.16 -.06 .33 ** .30 ** -.50 *** -.41 *** Attractiveness -.09 -.10 -.05 -.04 .11 .11 -.07 -.06 Expertise .06 .03 .03 .05 .01 .00 -.06 -.04 Perceived Leadership PLS .29 ** .42 *** -.05 .34 *** PLS - Inspiring .48 ** .03 .04 .03 PLS - Dominant -.01 .29 *** -.06 .24 *** PLS = Perceived Leadership Scale Model 1 Model 3Model 2 **p < .05. ***p < .01. Model 2 Model 1 Model 2 Model 3Model 1 Model 3 Model 3Model 1 Model 2 - 50 - Consistently, results show the predictive power of perceived leadership on disappointment (β = .29, p < .05), blame (β = .42, p < .00), and intention in wrongdoing (β = .34, p < .00). The results suggest that when a politician with higher perceived leadership is involved in a wrongdoing, the public feels more disappointed, more likely to blame this person, and more likely to believe this person was intentional in their wrongdoing. Further, perceived leadership did not influence people’s tendency to forgive (β = .05, n.s.), but the perception of trustworthiness (β = .33, p < .05) and general liking (β = .32, p < .00) were found to be predictors. The results of Model 3 show more determinant effects from the perception of a leader’s inspiring qualities (PL – inspiring) on disappointment (β = .48, p < .00) than the perception of a leader’s dominant and power qualities (PL – dominant) which, yield stronger effects on blame (β = .29, p < .00) and intention (β = .24, p < .00). These results suggest that when a leader is involved in a wrongdoing, people feel more disappointed by an inspiring leader, but put more blame on a dominant leader and interpret this leader’s mistake as purposeful. This study provides the first evidence demonstrating significant effects of perceived leadership on consumer moral judgments. These results suggest the importance of adding perceived leadership as a new endorser dimension, and further they differentiate the proposed construct from the traditional endorsers’ dimensions. Study 5 – Exploring Brand Mind Perception and Individual Differences This study explores whether consumers can perceive a brand as having a mind. This research postulates that individual differences found important in anthropomorphism can be applied to influence people’s tendency to perceive a brand’s mind. Thus, this - 51 - study aims to examine the three propositions P1, P2, and P3 about individual differences first introduced in Chapter 2. Further, this study examines brand associations correlated with brand mind perceptions, and explores whether brand mind perception influences consumers’ moral judgments, providing preliminary evidence that: 1) CEO association may influence brand mind perception, and 2) brand-mind perception matters in consumer moral judgment of a brand. Data, Measurement, Statistical Methodology. A survey was designed as a correlational study and conducted using Amazon mTurk panel for data collection. A total of 117 adults from the United States started the study, but only 104 (Mage = 33.5, 42% female) participants completed the study and were paid for their participation. To understand consumers’ tendency toward seeing a brand as having a mind, the questionnaire started with agree/disagree or yes/no closed questions. Participants were asked 1) whether they agree with the statement that “I sometimes imagine non-human things to have a human-like mind;” 2) whether they would answer “yes” or “no” to the question “Are there any brands or companies for which you can think of as having a mind similar to how you might think of another person having a mind?” and 3) whether they agree with the statement: “I sometimes think about a certain brand or company as having a mind similar to how I think about another person having a mind.” Individual Differences. Participants indicated their overall brand-mind perception tendency by two items (r = .85): “I can easily think of a company or brand that I would describe as having a mind” and “I think about the characteristics of a brand’s mind in a manner similar to how I think about the characteristics of a human’s mind” with five- point scales anchored by 1 = strongly disagree and 5 = strongly agree. - 52 - Participant demographic information was collected including age, gender, location, marital status, race, education, and religious beliefs. See the Appendix H for details regarding demographics. Further, participants responded to individual difference measures including interdependence of self-construal (2-item; Escalas & Bettman, 2005), loneliness (6-item; Gierveld & Tilburg, 2006), and need-for-cognition (5-item; Wood & Swait, 2002). First, the software SAS 9.4 was used to run a GLM with the brand-mind perception tendency as the dependent variable and the demographic variables as predictors. Second, the model was run again by adding individual difference (interdependence, loneliness, and need-for-cognition) as predictors. Brand Associations. To investigate what brand associations influence mind perception tendency, the survey asked the participants to name a brand they could perceive as having a mind, as well as a brand they could not perceive as having a mind. Participants then evaluated these two brands in terms of top-of-mind associations and brand personality as well as control variables regarding brand experiences. For top-of- mind association with a brand’s products, the survey asked participants to what degree they agree with the statement that “When I think of this brand, I will think about its products” anchored by 1 = strongly disagree and 5 = strongly agree. For other top-of- mind associations, the survey asked them the same questions but replaced the word “products” with “employees,” “leader/CEO,” “spokesperson (human or character),” and “logo.” To differentiate brand mind perception from other factors in influencing consumer attitudes, the control variables included brand personalities (e.g., being sincere, exciting, competent, sophisticated, and rugged, five-point Likert scales, (Aaker, 1997), as well as - 53 - brand familiarity, purchase, and positive and negative experiences by asking to what degree they agree or disagree to the statements: 1) I am very familiar with this brand; 2) I have purchased this brand’s products or services; 3) I have had some positive experiences with this brand; and 4) I have had some negative experiences with this brand, anchored by 1 = strongly disagree and 5 = strongly agree. SAS 9.4 was used to run logistic regression models with brand associations as predictors and a brand having a mind versus a brand not having a mind as the outcome while controlling for brand personalities and brand experiences. Exploring Moral Consequences. To capture the potential effects of brand mind on moral judgments, participants were asked to respond to questions about their moral judgment and action intentions toward the four nominated brands (no mind, yes mind, agency mind, or experience mind). Specifically, these moral questions explored four aspects by asking participants to what extent they agree with the statements anchored by 1 = strongly disagree and 5 = strongly agree. The first moral aspect was the positive moral intentions including helping (“I would feel good about doing something to help this brand”), appreciation (“if this brand does something to treat me well, I would feel very grateful to this brand”), and loyalty (“I’m loyal to this brand”). The second aspect explored how consumers would respond to a brand’s wrongdoings, such as forgiving (“If this brand apologizes for a harmful error, I will forgive it”), and punishment (“If this brand wrongs me, I would want it punished”). The last items about moral intention related to what consumers would perceive a brand would feel, such as recognition (“If others treat this brand well, I think this brand will recognize and appreciate it”). SAS 9.4 was used to run a mixed regression model with repeated measurements. This procedure - 54 - allowed using a categorical variable, a brand having a mind versus a brand having no mind (1, 0), as a predictor of consumer moral judgments and intentions while controlling for brand experience and brand personalities. Results. As expected, the results show that not all people perceive brands having minds (57% of participants indicated that they could). Compared to perceiving other non- human subjects (e.g., trees or animals) as having minds, participants did not find brand to be significantly different in terms of having a mind (χ2 = .106, p = .30). Individual Differences. The GLM model with demographic variables as predictors revealed that education and religious beliefs had significant effects on brand-mind perception (r = .85) tendency. See Table 6 for details. As exploratory research, the significance level was set at .10 for all analyses. Participants who have college or higher education indicated less brand-mind perception than participants with no college education (β = .66, p < .05). Further, participants who reported their religious beliefs as “spiritual” indicated higher brand-mind perceptions (β = .57, p < .10). The full GLM with both demographic variables and individual difference measures revealed that interdependent self-construal (r = .70) had significant positive effects on the tendency of seeing a brand having a mind (β = .34, p < .05). In addition, loneliness (α = .74) had positive effect (β = .62, p < .10) while need-for-cognition (α = .84) had negative effect (β = -.28, p < .10). These results suggest that the tendency to perceive a brand as having a mind varies among people and depends, at least in part, on their loneliness (P1), interdependence (P2), and need-for-cognition (P3). A person who is more interdependent or feels lonelier is more likely to see a brand or a company as having a mind. However, a person who has higher need-for-cognition is less likely to perceive a brand’s mind. - 55 - Table 6 Study 5: Results of GLM Model of Perception of a Brand as Having a Mind Age -0.01 -0.01 0.00 0.00 0.01 0.01 (0.01) (0.01) (0.01) (0.01) (0.01) (0.01) Gender Female -0.09 -0.09 0.13 -0.01 -0.03 -0.14 (0.25) (0.25) (0.26) (0.25) (0.24) (0.23) Location (Reference Group (Ref.): South) West -0.17 -0.12 0.25 -0.31 0.26 -0.33 (0.33) (0.32) (0.35) (0.32) (0.31) (0.29) Northeast -0.29 -0.11 0.01 -0.23 -0.37 0.13 (0.35) (0.34) (0.36) (0.35) (0.33) (0.31) Midwest -0.51 -0.39 -0.36 0.25 0.10 -0.18 (0.34) (0.32) (0.35) (0.33) (0.31) (0.29) Marital (Ref.: Divorced/Separated/Widowed) Married 0.62 0.79 * 0.65 0.92 * 0.47 0.61 (0.47) (0.46) (0.49) (0.47) (0.44) (0.42) Never-Married 0.50 0.72 0.52 0.80 0.49 0.67 (0.50) (0.49) (0.52) (0.50) (0.47) (0.45) Race (Ref.: Non-Hispanic White) Latino or Hispanic 0.22 0.12 0.36 0.25 0.45 0.31 (0.43) (0.41) (0.44) (0.42) (0.40) (0.37) Black or African 0.46 0.81 * 0.54 0.88 * 0.43 0.71 * (0.46) (0.45) (0.48) (0.46) (0.43) (0.41) Asian -0.03 -0.24 0.54 0.35 1.50 *** 1.21 ** (0.52) (0.50) (0.54) (0.51) (0.49) (0.45) Education (Ref.: College Education and Higher) No College Education 0.66 ** 0.64 ** 0.56 ** 0.54 ** 0.22 0.19 (0.26) (0.25) (0.27) (0.26) (0.24) (0.23) Religion (Ref.: Christian) Spiritual 0.95 * 1.07 * 0.61 0.67 1.16 ** 1.22 ** (0.57) (0.55) (0.59) (0.56) (0.53) (0.50) Others 0.24 0.36 0.34 0.36 -0.03 0.06 (0.30) (0.29) (0.31) (0.30) (0.28) (0.27) Individual Difference Implicity 0.00 0.00 -0.03 (0.06) (0.06) (0.06) Self-Construal-Interdependent 0.34 ** 0.22 0.26 ** (0.14) (0.14) (0.13) Loneliness 0.62 * 0.90 ** 0.96 *** (0.34) (0.34) (0.31) Need-For-Cognition -0.28 * -0.46 *** -0.22 (0.15) (0.15) (0.14) Control Variables: Segmentation Descriptors Intercept 2.24 * -0.51 2.20 * -0.60 1.17 -2.43 * (0.88) (1.36) (0.91) (1.39) (0.82) (1.23) Demographic Variables *p < .10. **p < .05. ***p < .01. Notes: Standard errors are listed in parentheses below the parameter estimates. Model 1 Model 2 Model 1 Model 2 Model 1 Model 2 T ble 1 Results: GLM Model of Perceptions of Mind, Agency Mind and Experience Mind Overall Mind Agency Mind Experience Mind - 56 - Brand Associations. To explore what factors associated with a brand would influence perception of brand mind, a logistic regression model was conducted with mind perception (yes, no) as the dependent variable. The results revealed that top-of-mind association with leader/CEO positively impacted the outcome of brand-mind perception (β = .31, p < .05). See Table 7 for details. This result provided confidence for the following studies to investigate the effects of CEO endorsers on brand-mind perception and consumer moral judgments. Further, top-of-mind association with brand logo positively impacted the outcome of brand-mind perception (β = .32, p < .05), and in contrast, top-of-mind association with purchase experience negatively influenced brand- mind perception (β = -.73, p < .01). Because “purchase” is a more specific thought about a brand while a “logo” is an abstract representation of a brand, the results imply that brand-mind perception tendency may positively (negatively) correlate with abstract (concrete) brand thoughts. This speculation is not empirically tested within the dissertation but is discussed as an avenue for future research. - 57 - Table 7 Study 5: Results of Logistic Model of Perception of a Brand as Having a Mind Consequences. The next analysis explores the relationships between perception of brand mind and potential moral consequences. Although no causal relationship could be vs. No Mind vs. No Mind vs. No Mind Familiar 0.16 0.25 0.23 0.32 0.22 0.24 (0.22) (0.24) (0.22) (0.25) (0.21) (0.22) Purchased -0.68 *** -0.73 *** -0.35 * -0.20 -0.41 ** -0.36 ** (0.21) (0.22) (0.19) (0.22) (0.17) (0.18) Positive Experience 1.03 *** 0.90 *** 0.69 *** 0.21 0.47 *** 0.20 (0.22) (0.27) (0.19) (0.25) (0.17) (0.23) Negative Experience -0.24 -0.15 -0.32 ** -0.21 -0.25 * -0.23 (0.15) (0.17) (0.15) (0.18) (0.14) (0.16) Brand Association Variables Top-of-mind (TOM): Product 0.05 0.03 0.02 -0.16 0.03 0.15 (0.19) (0.21) (0.20) (0.23) (0.17) (0.19) TOM: Employee 0.02 -0.05 0.08 0.02 0.04 0.09 (0.13) (0.14) (0.14) (0.16) (0.13) (0.14) TOM: CEO/Leaders 0.31 ** 0.25 0.36 ** 0.30 * 0.03 -0.04 (0.15) (0.16) (0.15) (0.16) (0.13) (0.14) TOM: Spokesperson or Characters -0.06 -0.03 -0.14 -0.22 0.13 0.15 (0.15) (0.16) (0.16) (0.18) (0.14) (0.15) TOM: Logo 0.36 ** 0.32 ** 0.36 ** 0.21 0.16 0.08 (0.15) (0.15) (0.15) (0.17) (0.13) (0.13) Brand Personality Variables Sincere -0.09 -0.04 -0.04 (0.23) (0.22) (0.20) Exciting 0.27 0.55 ** 0.73 *** (0.25) (0.25) (0.25) Competent -0.15 0.20 -0.21 (0.25) (0.24) (0.22) Sophisticated 0.45 * 0.42 * -0.08 (0.26) (0.25) (0.22) Rugged -0.21 -0.12 -0.39 ** (0.17) (0.16) (0.17) Intercept -3.26 ** -3.61 *** -3.36 *** -4.70 *** -1.46 -1.18 (1.27) (1.39) (1.29) (1.52) (1.17) (1.32) Brand Experience Variables *p < .10. **p < .05. ***p < .01. Notes: Standard errors are listed in parentheses below the parameter estimates. Model 1 Model 2Model 1 Model 2 Model 1 Model 2 Table 2 Results: Logistic Model of Perceptions of Brands with Mind, Agency Mind, and Experience Mind Overall Mind Agency Mind Experience Mind - 58 - established by a survey study, the data were examined to see if any significant correlations indicate potential effects from brand mind perception on moral judgments. The mixed regression model analyses revealed that after controlling for brand experience (familiarity, purchase, positive and negative experience) and brand personality (sincere, exciting, competent, sophisticated, and rugged), brand-mind perception had significant effects on positive intentions such as help (β = .47, p < .01), appreciation (β = .25, p < .01), and loyalty (β = .18, p < .01), as well as intentions after a brand’s wrongdoing such as forgiveness (β = .18, p < .05) and punishment (β = -.14, p < .10). These results suggest that brand-mind perceptions do influence consumer moral judgments, and such outcomes are beyond those of brand personalities and experiences that have been discussed in previous research (e.g., Aaker et al., 2004). Discussion. Although not all people perceive a brand as having a mind similar to a person having a mind, this study shows that people have similar likelihood of perceiving a brand as having a mind, as contrasted with anthropomorphizing other non- human subjects and identifying their mind. In addition, this likelihood varies depending on individual differences such as consumers’ social needs and cognitive styles such as need-for-cognition, supporting propositions P1, P2 and P3. This study also shows that brand association with a CEO can contribute to forming brand-mind perception. These results imply that placing a CEO into the spotlight of marketing communication may enhance perceptions of brand mind. Finally, the data provide early evidence that perception of brand mind may influence consumer attitudes and moral judgments of a brand, although the study only gives a picture for brand-mind perception without specific contexts. - 59 - Study 6 – 9: Investigating the Proposed Moderated Mediation Model of Consumer Moral Judgments The two exploratory studies provide important insights as well as confidence in investigation of the proposed conceptual model (see Figure 2). Specifically, Study 4 shows that perceived leadership is distinct from traditional endorser dimensions, especially in predicting consumer moral judgments; Study 5 finds that the concept of mind perception is applicable to a brand, at least to some extent for most individuals, and that brand mind perception may be a critical link to consumer moral judgment. This section with four studies will directly test the hypotheses in the conceptual model which predicts that perceived leadership induced by a CEO endorser influences brand mind perception, which in turn influences consumer moral judgments toward a brand. One important theory to demonstrate is that perception of a brand endorser can shift perception of a brand itself. Study 6 first compares CEO endorsers with conventional brand endorsers (e.g., athlete) and examines how a CEO endorser elicits perceived leadership and thus influences perception of a brand and enhances consumer attitudes; Studies 7, 8 and 9 investigate brand wrongdoing situations and find that perceived leadership alters perception of brand mind, which in turn determines moral judgments toward a brand. Study 6 - Perceived Leadership and Brand Attitudes This study examines whether consumers perceive higher leadership from a CEO rather than other commonly used endorsers such as athletes (H1). Before the - 60 - investigation of the effects on consumer moral judgment in a brand’s wrongdoing scenario, this study starts with examination of consumer attitudes in a general marketing communication setting. Thus, this study tests the prediction that perceived leadership induced by a CEO endorser will positively influence consumer perceptions and attitudes toward a brand; and, in particular, that featuring a CEO as a brand endorser in marketing communications will positively influence consumer perception of a brand as a leading brand in its industry (H2), as well as positively influence consumer attitudes toward a brand (H3), and perceived leadership mediates these relationships (H2 and H3). Design and Method. To contrast a CEO endorser with a conventional endorser, Study 6 employs a simple 2-condition (brand endorser: CEO vs. Athlete) between- subjects design. An athlete was chosen over a celebrity as a conventional endorser based on a pretest where participants (n = 108) indicated no difference between CEO-brand endorsers and athlete-brand endorsers in terms of trustworthiness, expertise, and attractiveness, but higher attractiveness and less expertise of celebrity-brand endorsers than CEO-brand endorsers (both ps < .01). College students participated in the study for course credit, and they were randomly assigned to evaluate a traveling mug print ad from a fictitious brand, Hudson. This brand name was borrowed from another study that had pretested the brand name and found it did not carry strong pre-existing associations (Chien, Cornwell, & Pappu, 2011). The print advertisement featured an endorser for Hudson, either the CEO of Hudson or an athlete, using the same name, “John Miller,” and a same photo. See Appendix B for stimuli example. A CEO as brand endorser was expected to induce higher perceived leadership than an athlete. - 61 - After viewing the print advertisement, participants immediately assessed their general attitudes toward the ads by the same items in previous studies. To capture their perception of the brand, Hudson, as a leading brand, they indicated to what extent they agree or disagree with the following statements about Hudson: 1) Hudson is a leader in its industry; 2) Hudson is a leading brand; 3) Hudson shows leadership. See Appendix H for further details on measures. As attention checks, participants then were asked: 1) whether they remembered John Miller in the advertisement; and 2) if so, who John Miller was (CEO, triathlete, or a movie star). Twenty-six out of a total of 83 participants failed attention checks thus were removed from the data, resulting in 57 participants in the final data analysis. This study was arranged as the last study in a lab data collection session which included a total of five independent studies. Thus, it is suspected that the working load may contribute to the result of a high number of participants failing the attention checks. Participants then rated “John Miller” by the 7-item perceived leadership (PL) scale including sub-dimensions PL – inspiring (4 items) and PL – dominant (3 items), as well as traditional endorser scales for trustworthiness (5 items), expertise (5 items), and attractiveness (5 items) as utilized in Studies 3 and 4. Results – Construct Validity. To provide further evidence supporting construct validity, the relations of PL with trustworthiness, expertise, and attractiveness were analyzed, yielding similar results as Study 3 and Study 4. Specifically, the data confirmed the convergence between PL (α = .872) and trustworthiness (α = .917; β = .410, p < .01) and expertise (α = .695; β = .232, p < .10). Further, PL showed no significant relationship - 62 - with attractiveness (α = .865), confirming the expected discriminant validity between these two constructs. Mediation Paths on Consumer Attitudes. First, a series of one-way ANOVAs were conducted to examine the difference in endorser perceptions of a CEO endorser versus athlete endorser. The results confirmed that participants perceived a CEO endorser as having significantly higher PL (M = 4.61) than an athlete endorser (M = 3.98; F(1,55) = 5.25, p < .05). Furthermore, a CEO endorser was rated significantly higher in both sub- dimensions, namely, PL – inspiring (α = .917; M = 4.53) and PL – dominant (α = .785; M = 4.71) than an athlete endorser (Minspiring = 3.93; F(1,55) = 4.12, p < .05; Mdominant = 4.06; F(1,55) = 4.92, p < .05). These results supported that a CEO endorser (vs. athlete endorser) induces higher PL, thus supporting H1. In addition, the analyses showed that the two conditions did not differ in trustworthiness (MCEO = 4.49, Mathlete = 3.94; F(1,55) = 3.71, n.s.), expertise (MCEO = 4.79, Mathlete = 4.45; F(1,55) = 1.33, n.s.) and attractiveness (MCEO = 4.26, Mathlete = 3.82; F(1,55) = 2.97, n.s.). It was expected that featuring a CEO rather than an athlete in advertising would lead to higher PL, which in turn influences consumers’ perception of the brand (i.e., “leading brand” perception, H2) and leads to positive consumer attitudes (i.e., brand ad attitude, H3). To examine this prediction, regression-based path analyses (Hayes, 2013; Preacher & Hayes, 2004) were conducted with 10,000 bootstrap samples to test the proposed mediated (via PL) model with leading brand perception (α = .860) or brand ad attitude (α = .954) as a dependent variable respectively (See Figure 3 for mediation model graphics). The analyses revealed a positive effect of featuring a CEO (vs. athlete) as an endorser on ad attitude mediated by PL (β = .471, SE = .249, 95% CI = [.093 to 1.089]). - 63 - The same mediation path (see Figure 3) emerged for enhancing the perception of being a leading brand (β = .352, SE = .193, 95% CI = [.054 to .835]) as an outcome of interest. Thus, hypotheses 2 and 3 were supported. Figure 3 Study 6: Simple Mediation Paths Alternative Mediation Paths. The traditional endorser perceptions such as trustworthiness, attractiveness, and expertise have been shown as important endorser- initiated factors to influence consumer attitudes toward a brand (Ohanian, 1990; Spry et al., 2011). The result of one-way ANOVA analysis suggested that except for the contribution of the perceived leadership measure, simply being a CEO or an athlete would not lead to different perceptions in the three traditional dimensions of expertise, trustworthiness and attractiveness. However, to further understand how the four perceptions (three existing and one proposed) would influence consumer attitudes together, regression-based path analyses (Hayes, 2013; Preacher & Hayes, 2004) were conducted with 10,000 bootstrap samples to test a few potential models (See Figure 4). - 64 - Figure 4 Study 6: Alternative Mediation Paths First, a parallel mediation model with four endorser perceptions as potential mediators (Model 3). The results showed that PL was the only mediating the effect of featuring a CEO (vs. an athlete) on brand ad attitudes (β = .269, SE = .214, 95% CI = [.003 to 886]). Second, a sequential mediation model with four endorser perceptions as potential mediators and PL positioned as the first (Model 4) yielded two significant mediation paths: 1) via PL (β = .286, SE = .222, 95% CI = [.004 to .908]); 2) via PL and trustworthiness (β = .237, SE = .176, 95% CI = [.019 to .819]). Furthermore, when a sequential model was conducted without PL as the first, the analyses yielded no significant paths. Note: black lines indicate significant paths - 65 - These results confirm that a CEO as a brand endorser is unique and brings new perceptions to consumers compared to traditionally studied endorsers such as athletes and celebrities. More importantly, these results together support the proposal that perceived leadership can function as a pivotal factor in influencing consumer attitudes beyond the traditional endorser characteristics, thus demonstrating the importance and necessity of introducing perceived leadership into marketing research. The next study investigates how perceived leadership influences consumer responses in a brand’s wrongdoing situation. Study 7 - Perceived Leadership and Consumer Moral Judgments The main purpose of Study 7 is to capture the effects of perceived leadership on consumer moral judgments toward a brand in a brand-wrongdoing situation. The proposed conceptual model (see Figure 2) predicts that brand mind perception is the critical link between perceived leadership and consumer moral judgments. Thus, the success of this study will pave the way to test brand mind perception as the critical mediator in subsequent studies. In addition, this study explores boundary conditions related to issue controllability (H8). Design and Method. Study 7 employed a 2 (brand endorser: CEO vs. athlete) x 2 (crisis controllability: high vs. low) between-subjects factorial design. Participants from the Amazon mTurk online panel took the study for a small cash incentive. They were randomly assigned to one of four conditions to respond to a brand’s print advertisement and then a news article about a product recall that served as the wrongdoing scenario. The first session of the study repeated the format of Study 6 by asking participants to evaluate - 66 - a print ad from a fictitious brand, Hudson. The same stimuli were used (print ads featuring either a CEO or an athlete as the endorser), and participants responded to the same questions about ad attitudes (3 items), leading brand perception (3 items), as well as two attention checks that asked whether they remembered John Miller in the advertisement, and who John Miller is (CEO, triathlete, or a movie star). Twenty-nine out of 165 participants failed the attention checks. They were removed from the data set with one additional participant who took the study twice, thus resulting in 135 (Mage = 36.40, 45.9% female) samples for the final data analysis. As in Study 6, participants evaluated PL (7 items including 4 items for PL – inspiring, and 3 items for PL – dominant), and traditional endorser scale including trustworthiness (5 items), expertise (5 items), and attractiveness (5 items). Participants indicated leading brand perception (3 items) of Hudson as well as their attitudes toward the advertisement (3 items) by the same measures in Study 6. See Appendix H for measure summary. The second section of the study presented participants a fictitious CNN news article about Hudson’s product recall over poisonous material being used in producing the product. This news story was fashioned after available announcements of a similar kind and clearly identified the product recalled was the one in the print ad that the participants had evaluated earlier. See Appendix C for stimuli examples. To manipulate the issue controllability, the news article either suggested that the contamination happened in Hudson’s own production facility (high controllability) or a supplier’s production facility (low controllability). In general, participants were expected to form more stringent moral judgments toward a brand when its CEO was in the spotlight of marketing communications than when a traditional endorser (i.e., an athlete) was. - 67 - Further, the expectation was that the endorser effects would disappear when a crisis was highly controllable by the company. This expectation is because the high controllability of a crisis leads to high attribution of responsibility (Russell, 1982), leaving little room for other factors to have an impact on consumer moral judgments. To measure moral judgments, the study asked participants to indicate to what extent they agreed with the following statements: 1) “I will blame Hudson, the brand, for the poisonous materials used in its products;” 2) “I believe Hudson, the brand, is responsible for the poisonous materials in its products;” 3) “It is Hudson, the brand’s fault for using the poisonous materials in its products” (1 = strongly disagree; 7 = strongly agree; adapted from De Matos & Rossi, 2007). These three items formed the blame score (α = .936). In addition, punishment intention was measured by asking “how much do you think the brand deserves punishment?” (1 = not at all; 7 = very much), as well as forgiveness by asking “how likely would you forgive the brand if the brand apologized?” (1 = very unlikely; 7 = very likely). See Appendix H for measure summary. For manipulation check of the issue controllability (α = .682), participants were asked to indicate “to what extent you agree or disagree with the statements: 1) the causes of poisonous materials were controlled by the company; 2) the causes of poisonous materials were always present; 3) the product crisis happened in a process which was under control of the company” (1 = strongly disagree; 7 = strongly agree; adapted from McAuley et al., 1992). Results – Endorser Perceptions and Ad Attitudes. To test whether a CEO (vs. an athlete) as an endorser would differ in PL and traditional endorser perceptions, a series of ANOVAs were conducted. The results confirmed that a CEO endorser was perceived - 68 - higher in PL (α = .932; M = 5.47) than an athlete endorser (M = 4.98; F(1,133) = 6.22, p = .01), but both were at the same levels in terms of trustworthiness (α = .951), attractiveness (α = .892), and expertise (α = .931). Further, the path analyses (Hayes, 2013; Preacher & Hayes, 2004) with 10,000 bootstrap samples successfully replicated the mediation paths on consumer attitudes found in Study 6. The results showed that featuring a CEO as a brand endorser would positively influence “leading brand” perception (α = .926; β = .272, SE = .103, 95% CI = [.090 to .502]) as well as positively influence ad attitudes via PL (α = .968; β = .274, SE = .109, 95% CI = [.092 to .538]). PL and Brand Moral Judgments. Before the effects on consumer moral judgments were examined, the manipulation of controllability was checked. The one-way ANOVA confirmed the treatment was successful (α = .682; Mhigh = 5.18; Mlow = 4.34; F(1,133) = 16.59, p < .01). Then, two-way ANOVAs were conducted with endorser (CEO vs. athlete) and controllability (high vs. low) as independent variables and blame (α = .936) as a dependent variable. The analyses yielded a main effect of controllability (F(1,131) = 11.89, p < .01), a marginal effect of endorser (CEO vs. athlete; F(1,131) = 3.50, p = .06), and an interactive effect of controllability and endorser (F(1,131) = 4.87, p < .05). See Table 8 for means and Figure 5 for effect graphics. Table 8 Study 7: Table of Means (+/- Standard Error) of Moral Judgments Blame Punish Forgive Blame Punish Forgive High Controllability 6.00 5.92 3.43 6.07 5.67 3.33 (.19) (.22) (.24) (.21) (.24) (.27) Low Controllability 5.75 5.57 3.66 4.94 4.94 3.97 (.20) (.23) (.25) (.20) (.23) (.26) CEO Athlete - 69 - Figure 5 Study 7: Effects on Moral Judgements of a Brand’s Wrongdoing The results show that, in general, participants blamed the brand more when the controllability of material contamination was high (M = 5.35) rather than low (M = 6.03). This is expected because a company is held accountable for wrongdoings in the process under its control (Russell, 1992). However, when the controllability was low, a CEO endorser led to higher blame (M = 5.75) to the brand than an athlete endorser (M = 4.94; 3.50 4.00 4.50 5.00 5.50 6.00 6.50 High Controllability Low Controllability Moral Judgement: Blame CEO Athlete 3.50 4.00 4.50 5.00 5.50 6.00 6.50 High Controllability Low Controllability Moral Judgement: Punish CEO Athlete 1.50 2.00 2.50 3.00 3.50 4.00 4.50 High Controllability Low Controllability Moral Judgement: Forgive CEO Athlete - 70 - F(1,66) = 6.45, p = .01); as expected, this effect disappeared for conditions with high controllability, thus, supporting H8. Similar results emerged for punishment as the outcome of interest. A two-way ANOVA yielded a main effect of controllability on punishment (F(1,131) = 5.4, p < .05). Mean differences indicate that participants believed the brand deserved more punishment when the controllability was high (M = 5.81) than low (M = 5.26). Although the interactive effect on punishment was not significant, a similar pattern to that found in the analyses of blame appeared. Specifically, for low controllability conditions, participants indicated higher intention to punish the brand when a CEO (M = 5.57) was the endorser than when an athlete was (M = 4.94; F(1,66) = 3.91, p = .05). Again, there was no difference for high controllability conditions, suggesting a ceiling effect. When the incident was so blame worthy as being within the control of the company, there is reduced variance in blame and punishment. In addition, although no significant results were found for forgiveness as the outcome of interest, the main effect of controllability on forgiveness trended in the expected direction (F(1,131) = 2.80, p = .096), suggesting participants are less forgiving when controllability was high (M = 3.39) than low (M = 3.81), as expected by attribution theory (Russell, 1982). Perceived leadership is expected to play a critical role in determining moral judgments, as demonstrated by a moderated (by controllability) mediation (by perceived leadership) model of endorser (CEO vs. athlete) on brand moral judgments. Regression- based path analyses (Hayes, 2013; Preacher & Hayes, 2004) were conducted with 10,000 bootstrap samples to test the model. To compare PL with three traditional endorser measures, the Process procedure for SPSS - Model 14 (Hayes, 2013) were used and - 71 - included all four scales as parallel competing mediators. See Figure 6 for the path graphic. The analyses confirmed the only moderated mediation path is via PL. Specifically, for low controllability, PL mediated effects of CEO endorser on blame (β = .288, SE = .174, 95% CI = [.040 to .751]); however, for high controllability, the mediation effect of PL disappeared (95% CI = [-.420 to .027]). See Table 9 for the path analysis results. Figure 6 Study 7: Moderated Mediation Paths of CEO-Endorser Effect on Moral Judgment Note: black lines indicate significant paths - 72 - Table 9 Study 7: Result of Moderated Mediation Paths by Mediators Importantly, this study provides evidence to differentiate PL from traditional endorser perceptions by its predictive power in consumer moral judgments. Further, these results capture the relationship between perceived leadership and consumer moral judgment toward a brand, thus moving the research forward to the next step to investigate whether brand mind perception is critical for this relationship. Study 8 - Brand Mind Perception as the Link to Moral Judgments Previous studies show that when a brand is involved in wrongdoing, perceived leadership induced by a CEO endorser escalates consumers’ blame of the brand. Building on this finding, the proposed conceptual model (see Figure 2) predicts that perception of brand mind is the crucial link between perceived leadership and moral consequences. The current study is aimed to test this proposal and the related hypothesis, H5. Mediator Crisis Controllability Effect Boot SE BootLLCI BootULCI Perceived Leadership Low .2883 .1737 .0403 .7511 Perceived Leadership High -.0738 .1104 -.4196 .0265 Attractiveness Low .0089 .0491 -.0498 .1748 Attractiveness High .0008 .0279 -.0515 .0695 Trustworthiness Low -.0481 .1125 -.4488 .0652 Trustworthiness High .0229 .0699 -.0524 .2730 Expertise Low -.0894 .1768 -.5405 .1983 Expertise High -.0009 .1070 -.2343 .2076 Conditional indirect effect of CEO endorser (vs. athlete endorser) on blame at values of the moderator of crisis controllability - 73 - As well, this study extends the work thus far to include two new aspects. First, this study contrasts advertising featuring a CEO with advertising featuring an employee as endorsers. It is a common practice for a brand to have an employee as a brand ambassador in communications (Morhart et al., 2009). Findings comparing CEOs with employees will provide useful managerial implications since both CEOs and employees are possible choices from within the firm, typically at a minimum cost. Utilizing a CEO and an employee also offers further control of perceived areas of expertise since both types of endorsers should be knowledgeable about their firm and its products. Second, the previous studies consistently examined the consequences of using a brand’s own CEO as a brand endorser. It is the case that brands are at time endorsed by a CEO from another company and in doing so, bring their status to an advertisement. For example, Warren Buffet, the CEO of Berkshire Hathaway, has allowed his face to appear on all cans of Cherry Coke sold in China (Pham, 2017). This practice suggests that perceived leadership is a characteristic that can be utilized by a brand without stemming directly from the company advertising. With this concept in mind, the current study aims to test whether a CEO’s relation to a brand (i.e., a brand’s own CEO as an endorser, or a CEO from another brand) would make a difference in consumers’ moral judgment toward a brand’s wrongdoing (H7). Design and Method. Study 8 employed a 2 (brand endorser: CEO vs. employee) x 2 (brand related: yes vs. no) between-subjects factorial design. Participants (n = 200, Mage = 36.72, 45% female) from the Amazon mTurk online panel took the study for a small cash incentive. As in Study 7, participants were randomly assigned to one of four conditions to respond to a brand’s print advertisement and then a news article about - 74 - product recall; however, given the ceiling effects observed in Study 7, the wrongdoing was controllable in nature. Participants were first presented with a print ad about cough and flu medicine from a fictitious brand, Janssen, and the print ad featured either a CEO or an employee, Harry Watson, as an endorser. Further, Harry Watson was described either from Janssen, a medicine company (brand related: yes), or from another fictitious brand Avevo, a technology company (brand related: no). See Appendix D for stimuli design. After viewing the print advertisement, participants responded to the same questions as in previous studies about ad attitudes (3 items). As in previous studies, participants evaluated the endorser in terms of perceived leadership (PL; 7-item including 4 items for PL – inspiring and 3 items for PL – dominant), and traditional endorser characteristics of trustworthiness (5 items), expertise (5 items), and attractiveness (5 items). Perception of brand mind was measured by two adapted items (Morewedge, Preston, & Wegner, 2007): 1) “Please indicate the degree to which Janssen appears to have a mind of its own” (1 = definitely has no mind; 7 = definitely has a mind); 2) “Please indicate to what extent you agree or disagree with the statement that Janssen has mental capabilities (e.g., thinking, making plans, communication)” (1 = strongly disagree; 7 = strongly agree). See Appendix H for further measure detail. The second section of the study presented a mock news article from the Washington Post about Janssen’s product recall due to its serious and unexpected side effects. The product recalled was the same product as had been featured in the print ad that participants had evaluated. See Appendix E for the news stimuli. After reading the news article, participants responded to questions about their moral judgments toward - 75 - Janssen including blame (3 items), punish (1 item), and forgiveness (1 item) as in Study 7. Results – Endorser Perceptions. First, a series of two-way ANOVAs were conducted to explore how a CEO and an employee as a brand endorser would differ in perceived leadership and traditional endorser perceptions. As well, the analysis explores whether a brand endorser related to the brand (yes vs. no) would influence the dependent variables of interest. The results yielded a consistent main effect of CEO vs. employee on PL and other endorser perceptions, but no main effects of brand relationship (yes vs. no) and no interaction effects. Specifically, a CEO endorser was viewed as higher in PL (α = .929; M = 5.13) than an employee endorser (M = 4.48; F(1,196) = 15.57, p < .01), including both PL sub-dimensions (PL – inspiring: α = .902; MCEO = 5.04; Memployee = 4.61; F(1,196) = 6.20, p < .05; PL – dominant: α = .894; MCEO = 5.25; Memployee = 4.32; F(1,196) = 28.56, p < .01), across brand relationship conditions. Furthermore, the results showed that a CEO endorser was perceived as more attractive (α = .850; MCEO = 4.35; Memployee = 3.95; F(1,196) = 7.68, p < .01) and expert (α = .927; MCEO = 5.33; Memployee = 4.63; F(1,196) = 15.26, p < .01) but at the same level in trustworthiness (α = .945; p = .23), again regardless of brand relationship conditions. It is not surprising that a CEO is perceived as having more expertise than an employee, given the fact that being a CEO usually has greater responsibilities, requires more education, training, skills and experience than being an ordinary employee. The result of attractiveness is, however, interesting. Given that the visual was controlled across conditions, the same person’s picture was presented to participants, the findings suggest that a person’s profession (or status) can make a person physically more attractive. This - 76 - finding can be explained in part by evolutionary psychology that posits psychological traits are adaptions, functional products of natural selection (Buss, 1991). CEOs as leaders may communicate higher survival or success probabilities than employees because of their higher power and wealth status, additional resources and skills. Ad Attitudes. Since CEO vs. employee conditions differed on PL as well as expertise and attractiveness, the effect paths on consumer attitudes toward brand ads were expected to be complex. It was expected that conditions using CEO or employee endorsers would influence consumer attitudes not only via PL but also via additional routes involving attractiveness and expertise. Therefore, two path analyses were conducted (both using path analyses 9, Hayes, 2013; Preacher & Hayes, 2004; with 10,000 bootstrap samples) to find evidence supporting the predictions. First, a simple mediation path was conducted with CEO/employee as the X factor, ad attitudes (α = .971) as outcome the Y factor, and PL as mediator. The analysis confirmed the mediation path via PL (β = .514, SE = .138, 95% CI = [.270 to .792]). Second, an additional mediation path analysis was conducted, this one adding PL, as well as attractiveness, expertise, and trustworthiness as sequential mediators. The results indicated three mediation paths on ad attitudes: (1) via PL => Expertise (β = .169, SE = .102, 95% CI = [.002 to .419]); (2) via PL => Expertise => Trustworthiness (β = .248, SE = .101, 95% CI = [.095 to .500]); and (3) via PL => Attractiveness => Trustworthiness (β = .009, SE = .007, 95% CI = [.0002 to .032]). The results of this study are consistent with previous research, showing that endorser perceptions including expertise, attractiveness, and trustworthiness influence consumer attitudes toward a brand and its offering (Ohanian, 1991). In addition, these results reflect the discussion in the scale development - 77 - section about the relationships between perceived leadership and expertise and trustworthiness, as well as a discussion in this study (see results of endorser perceptions) that evolutionary psychology provides an explanation about why CEOs are perceived higher in attractiveness than employees. Furthermore, the current study supports the proposal of perceived leadership as a new endorser perception by offering additional evidence that leadership perceptions stemmed from a CEO endorser positively influence consumer attitudes. Brand Mind Perception and Brand Moral Judgments. To investigate the moderating effects of the brand relationship of the endorser, a two-way ANOVA was conducted with CEO endorser (vs. employee) and brand relationship (yes or no) as independent variables and perception of brand mind as a dependent variable. The results reveal a significant interaction (F(1,196) = 27.49, p < .01). Specifically, featuring a brand’s own CEO led to significantly higher perception of brand mind (r = .929; M = 4.53) in contrast to featuring a brand’s employee (M = 3.45; F(1,97) = 7.17, p < .01). However, the effect disappeared when a CEO or employee was not related to the brand being advertised. See Figure 7 for result graphic. This result demonstrates the moderation effects of brand relationship on perception of brand mind. - 78 - Figure 7 Study 8: Interactive Effect on Perception of Brand Mind For the analysis of moral judgments, there were two main tasks for Study 8: 1) building the mediation path via brand mind perception between PL and moral judgments; 2) testing the moderation effects from the brand relationship of the endorser. Perceived leadership initiated by a CEO endorser is expected to positively influence the perception of brand mind (H4), which in turn is expected to alter consumer moral judgments of a brand (H5). Further, it is hypothesized that when a CEO is related to the brand the effect will be signification, but will disappear when a CEO is not related to the brand being advertised (H7). Please see Figure 8 for the model graphic representation of the expected mediated moderation model. - 79 - Figure 8 Study 8: Moderated Mediation Model of PL Effects on Moral Judgment Earlier analyses on endorser perceptions had confirmed the positive main effect of CEO endorser on perceived leadership (Studies 6 and 7). Thus, the following analysis focuses on the moderated mediation paths starting from PL. Regression based path analysis (Hayes, 2013; Preacher & Hayes, 2004) was conducted with 10,000 bootstrap samples to directly test the moderated mediation paths with blame as the outcome of interest while controlling for trustworthiness, expertise, and attractiveness. The results confirmed the prediction that for conditions where the endorser was brand-related, brand mind perception mediated the effects of PL on blame (α = .821; β = .030, SE = .021, 95% CI = [.001 to .091]); however, no significant mediation effects were found when the endorser held no relation to the brand (95% CI = [-.077 to .006]). Similarly, when using “forgiveness” as the outcome of interest, the analysis again yielded the moderation effect of brand relationship on the mediation path. For brand-related conditions, brand-mind perception mediated the effect of PL on forgiveness (β = -.059, SE = .042, 95% CI = [- .177 to -.001]); again, this effect disappeared for conditions where the endorser was not - 80 - related to the brand being advertised (95% CI = [-.013 to .139]). Thus, the moderated (H7) mediation (H4 and H5) effects suggested were supported. Together, these results show that in a situation where a brand has done wrong of some kind, higher leadership perceptions initiated by a CEO endorser lead to higher perceptions of brand mind, which in turn influences consumer moral judgment toward a brand (more blame and less forgiveness). However, this effect only occurs when the CEO endorser is related to the brand advertised. CEOs and employees from another brand will be less effective as brand endorsers in influencing brand perceptions; on the other hand, these more distant endorsers are also “safer” in situations when brands need forgiveness for wrong doings--they will be judged less harshly with a distant endorser. The next study will examine leadership perception with consideration of the two sub-dimensions (inspiring vs dominant) and thus further examine how an inspiring CEO vs. a dominant CEO may influence consumer moral judgments. Study 9 – Considering Inspiring and Dominant Dimensions of Leadership The main purpose of this study is to investigate whether manipulating perceived leadership in terms of the two sub-dimensions, inspiring and dominant leadership, will result in distinct consumer moral judgments toward a brand’s wrongdoing. It is expected that dominant leadership leads to more rigid moral judgments (more blame and punishment) than inspiring leadership, which should lead to more flexibility (more willing to forgive). The dominant leader is associated with power and authority in decision-making. It is argued that these characteristics signal ultimate responsibility for - 81 - actions (Gray & Wegner, 2009). Further, perception of the experience aspect of the brand mind (H6) is hypothesized as the underlying driving force for these opposite results. Design and Method. Study 9 used a 4 (endorser type: employee, CEO-general, CEO-inspiring, CEO-dominant) x 2 (issue controllability: high vs. low) between-subjects factorial design. The conditions of employee versus general CEOs as brand endorsers were used to produce high versus low leadership perceptions, and the conditions of inspiring CEO versus dominant CEO were intended to induce specific leadership perceptions as inspiring leadership versus dominant leadership. More details will be discussed in the following paragraphs. Participants were adults from the Amazon mTurk subject pool and were paid a small incentive for completing the study. Study 9 followed a similar procedure as in previous studies where participants were first exposed to leadership treatments, and then responded to a news article about a brand’s wrongdoing. However, the leadership treatment materials in this study were meaningfully altered from the print advertisement used in previous studies (which featured either a CEO endorser or an athlete/employee endorser). To provide participants direct information about endorsers, the copy of the ad addressed the characteristics of the leader in the ad. This change allowed to manipulate different perceived leadership perceptions in Study 9. Participants first read a brief paragraph about a fictitious brand, Janssen, as utilized previously. Janssen was described as an American pharmaceutical company, introducing a new cold and flu relief medicine, and Harry Watson (either the CEO or an employee) from Janssen would work as the endorser for the new product. To manipulate perceived leadership, a brief introduction about the endorser was provided to participants. - 82 - The introduction described Harry Watson as either the CEO (high leadership) or an employee of Janssen (low leadership). In addition, to manipulate leadership perceptions as “inspiring” or “dominant,” two more conditions were added where the perceived leadership scale items were used to directly describe types of CEOs. Specifically, for “inspiring” leadership conditions, Harry Watson was described as a model leader well known as “responsible, courageous, and effective;” for “dominant” leadership conditions, Harry Watson was described as a leader known as “powerful and authoritative.” See Appendix F for details of stimuli design. After reading the endorser introduction, participants were asked to answer an attention check question and then manipulation check questions. The attention check asked participants to indicate “who Harry Watson is: 1) a CEO; 2) a celebrity; 3) an employee; and 4) I don't remember.” Among 400 participants, 32 of them failed the attention check and were thus removed from the data set, resulting in 368 respondents (Mage = 36.09, 46.7% female) in the final data analyses Single title items from perceived leadership dimensions were used for manipulation check, as well as single title items from the traditional endorser perceptions for control variables. Specifically, participants were asked to “indicate how much you believe in general Harry Watson is _____?” (the blank space was filled with “inspiring” “dominant” “trustworthy” “expert” or “attractive” respectively; anchored on 1 = not at all, 7 = very much). Participants responded to the same questions about perception of brand mind (2 items) as in Study 8. In addition, two specific mind measures, experience mind and agency mind, were measured by six items adapted from previous research in mind - 83 - perception (Gray et al., 2007). See Appendix H for measure summary. Specifically, for brand agency mind, three items included: 1) “To what degree can Janssen make plans and work toward goals?” 2) “To what degree is Janssen capable of thinking?” 3) “To what degree is Janssen capable of making decisions?” For brand experience mind, three items included: 1) “To what degree is Janssen capable of having experiences and being aware of things?” 2) “To what degree can Janssen feel joy?” and 3) “To what degree can Janssen experience embarrassment?” All items were anchored with the wording 1 = not at all; 7 = very much. Participants were then provided with a news article about Janssen’s product recall of the new medicine due to unexpected side effects. As in Study 7, issue controllability (high vs. low) was manipulated by the news article to suggest that the contamination happened either in Janssen’s own production facility (controllability high) or a supplier’s production facility (controllability low). See Appendix G for stimuli examples. Adding the crisis controllability allows this study to test the moderated (by crisis controllability) and mediation (by brand mind perception) paths in the proposed model all together. Participants indicated their moral judgments of the brand including blame (3 items), punishment (1 item), and forgiveness (1 item), all on a 7-point Likert scale. The same items (3 items) utilized as in Study 7 as a manipulation check for crisis controllability. Results – Leadership Perception Manipulation Check. A series of one-way ANOVAs with leadership manipulations (employee, CEO-general, CEO-inspiring, CEO- dominant) as independent variables and the single-item inspiring and dominant scores as dependent variables were run to test whether perceived leadership (employee vs. CEO) as well as sub-dimensions (inspiring vs. dominant) manipulations would work as expected. - 84 - The results revealed the significant differences of between the inspiring leadership endorser (F(3,364) = 7.65, p < .01), dominant endorser (F(3,364) = 30.76, p < .01), and overall PL score (F(3,364) = 18.03, p < .01) among the four endorsers. Tukey post hoc tests showed that participants rated a general CEO (M = 5.31) as well as an inspiring CEO (M = 5.56) and a dominant CEO (M = 5.70) significantly higher in PL than an employee as an endorser (M = 4.64; all ps <.01). In addition, an inspiring CEO was rated higher in inspiring score (M = 5.74) than a dominant CEO (M = 5.23; p <.05), a general CEO (M = 5.25; p =.05) or an employee (M = 4.83; p <.01). In contrast, a dominant CEO scored the highest on the dominant score (M = 6.28) compared to an inspiring CEO (M = 5.38; p <.01), a general CEO (M = 5.36; p <.01) or an employee (M = 4.45; p <.01). These results confirm a successful manipulation of PL (high vs. low) for an employee vs. a general CEO as a brand endorser, as well as a successful manipulation of PL sub- dimensions (inspiring vs. dominant) by describing a CEO as more courageous and effective (an inspiring CEO) or more powerful and authoritative (a dominant CEO). See Table 10 for means and Figure 9 for mean comparison graphic representation. Table 10 - 85 - Figure 9 Study 9: Perceived Leadership Scores Perceived Leadership (PL) PL - Inspiring PL - Dominant - 86 - One-way ANOVAs with Tukey post hoc tests on trustworthiness, expertise, and attractiveness as potential control variables revealed no significant difference of perceived trustworthiness and attractiveness among the four endorsers. However, the endorsers differed in perceptions of expertise (F(3,364) = 10.52, p < .01). Specifically, an employee endorser was perceived as less of an expert (M = 4.84) than all three CEOs (Mgeneral = 5.60, Minspiring = 5.72, Mdominant = 5.80, all ps < .01); but there was no difference among three CEOs. Thus, expertise was kept as a control variable in later analyses comparing high and low perceived leadership conditions (i.e., employee and CEO endorser conditions). Brand Mind Perceptions. This study aimed to replicate the result in Study 8 that perceived leadership of a brand endorser positively influences perception of brand mind. A GLM analysis was conducted using the categorical high vs. low PL (i.e., CEO-general vs. employee endorser) as an independent variable and perception of brand mind as a dependent variable while controlling for expertise. The results confirmed that higher PL perception led to higher perception of brand mind (r = .918; Mhigh = 3.99, Mlow = 2.99, F(1,178) = 9.80, p < .01). In addition, comparing perceptions of brand mind among three high PL (i.e., PL general, PL – inspiring, and PL – dominant) conditions found no differences (F(2, 279) = .47, p = .62). The next analysis examined the three high PL (manipulated by CEOs) conditions on manipulation of perceived leadership as inspiring or dominant would accordingly change perceptions of experience or agency in mind perception. Inspiring leadership and dominant leadership were not expected to differ in agency mind because the abilities to think, plan, and make decisions are foundational to typical leaders and thus are equally - 87 - expected. However, when contrasted with an inspiring leader, a dominant leader is expected to result in lower experience mind perception, which highlights capabilities of having experiences and being aware of things, such as pride or embarrassment; this result is because an individual’s dominant and powerful attributes generally have a negative relationship with experience in mind perception (e.g., God; Gray et al., 2007; Gray & Wegner, 2009). A one-way ANOVA with all three PL groups was conducted, and the result confirmed the predictions. Specifically, no difference was found in brand agency mind among high PL groups (α = .880; F(2,279) = .26, p = .77), but a significant difference in brand experience mind (α = .913; F(2,279) = 3.48, p < .05). Further, Tukey post hoc analysis showed that the difference in brand experience mind was driven by the gap between inspiring leadership (M = 3.93) and dominant leadership (M = 3.23, p < .05). See Figure 10 for mean comparison graphics, which also included the low PL group for reference. - 88 - Figure 10 Study 9: Perceptions of Brand Agency and Experience Minds by Perceived Leadership (PL) Conditions Brand Moral Consequences. First, the manipulation check for issue controllability (high vs. low) confirmed the treatment was successful (α = .754; Mhigh = 3.00 3.40 3.80 4.20 4.60 5.00 Low PL (Employee) High PL (General CEO) PL-Inspiring (Inspiring CEO) PL-Dominant (Dominant CEO) Perception of Agency Mind 2.00 2.40 2.80 3.20 3.60 4.00 4.40 Low PL (Employee) High PL (General CEO) PL-Inspiring (Inspiring CEO) PL-Dominant (Dominant CEO) Perception of Experience Mind p < .05 - 89 - 5.25; Mlow = 4.10, F(1,366) = 85.56, p < .01). Next two sets of analyses on brand moral judgments were made. The first set of analyses compared the high vs. low PL groups (i.e., employee vs. general CEO endorser). The results were expected to replicate the findings of Study 7, namely that higher PL results in more severe moral judgments, and issue controllability moderates this effect. In addition, the findings here are also expected to replicate the results in Study 8 that perception of brand mind mediates the effects between PL and moral judgments. Therefore, this set of analyses would provide a holistic picture by testing the mediation and moderation effects together. The second set of data analyses compared and contrasted the moral judgments across CEOs to test how manipulating PL based on the two sub-dimensions would influence consumers’ moral judgments. The results were expected to show that consumers would blame less and be more forgiving to an inspiring leader than a dominant leader, and that this effect is argued to be explained by the brand experience aspect of the brand mind (H6). Data Analysis Set 1. A GLM was operated with PL high vs. low (i.e., CEO vs. employee) and controllability (high vs. low) as independent variables and blame (α = .947) as a dependent variable while controlling for expertise. The analyses yielded a main effect of controllability (F(1,176) = 38.42, p < .01), and an interactive effect of controllability and PL (F(1, 176) = 5.21, p < .05). Specifically, the results showed that participants blamed a brand more when the controllability of material contamination was high (M = 5.93) instead of low (M = 4.79). Further, when the controllability was low, high PL led to higher blame (M = 5.16) to a brand than low PL (M = 4.34, p < .05). This effect disappeared for conditions with high controllability as occurred in Study 7. - 90 - To test the moderation as well as mediation paths, regression based path analyses (Hayes, 2013; Preacher & Hayes, 2004) were conducted with 10,000 bootstrap samples to test the moderated (by controllability) mediation (via brand-mind perception) paths with the categorical variable PL (high vs. low) as the independent variable and blame as the outcome of interest, while controlling for expertise. The results confirmed predictions that for low issue controllability conditions, brand-mind perception mediated the effects of PL on blame (β = .229, SE = .117, 95% CI = [.046 to .517]); however, no significant mediation effects are found for high issue controllability conditions (95% CI = [-.226 to .023]). Thus, the moderated mediation paths were supported (H4, H5, and H8). All these results support the general prediction that brand mind perception plays a critical role in determining moral judgments. See Figure 11 for path graphic, and Table 11 for path analysis results. Figure 11 Study 9: Moderated Mediation Model of PL Effects on Moral Judgment - 91 - Table 11 Study 9: Results of Moderated Mediation Path Analysis Data Analysis Set 2. The main task of this set of analyses was to compare and contrast different leadership groups in terms of their effects on brand moral judgments. Because the boundary effects of high issue controllability had been established, the low issue controllability conditions were used for this set of analyses. First, one-way ANOVAs showed that the CEOs (i.e., general leadership, inspiring leadership, and dominant leadership) significantly differed in term of blame (F(2, 144) = 3.60, p < .05), punishment (F(2, 144) = 3.35, p < .05), and forgiveness (F(2, 144) = 6.01, p < .01). Tukey post hoc tests revealed that these differences were driven by the gaps between inspiring leadership and dominant leadership groups. Specifically, dominant leadership led to higher brand blame (M = 5.60) and punish (M = 5.25) than inspiring leadership (Mblame = 4.82, p < .05; Mpunish = 4.48, p < .05); but no differences were found when compared with the general leadership conditions. In addition, inspiring leadership led to higher forgiving (M = 4.88) than dominant leadership (M = 3.93, p < .01) as well as general leadership (M = 3.94, p < .01). These results suggested that participants were more likely to give stringent moral judgments toward a dominant leader but less likely to do so toward an inspiring leader. See Figure 12 for comparison graphics. Mediator Crisis Controllability Effect Boot SE BootLLCI BootULCI Brand Mind Perception Low .2289 .1173 .0458 .5166 Brand Mind Perception High -.0723 .0618 -.2257 .0227 Conditional indirect effect of perceived leadership (high vs. low) on blame at values of the moderator of crisis controllability - 92 - Figure 12 Study 9: General vs. Inspiring vs. Dominant Leaderships on Moral Judgments - 93 - Brand-mind perception and brand experience aspect of the brand mind in particular were hypothesized to play a critical role in this mediation process (H5 and H6). Thus, the data among three PL groups (i.e., PL – general, PL – inspiring, and PL – dominant) were collapsed for a multiple regression analysis to test whether brand mind, brand experience mind, and brand agency mind would predict moral judgments toward a brand. Blame, punishment, and forgiveness were used as outcome variables, respectively. See Table 12 for the result summary. Consistently, results confirmed the predictive power of brand mind perception on blame (β = .29, p < .05) and punishment (β = .38, p < .00), but negatively on forgiveness (β = -.34, p < .00), supporting H5. In contrast, the models showed that brand experience mind predicted blame (β = -.28, p < .05) and punishment (β = -.27, p < .05), but also the positive effects on forgiveness (β = .24, p < .05), supporting H6. These results show that consumers blame a brand more for a brand’s wrongdoing and are less likely to forgive this brand when consumers perceive higher brand mind. However, higher perception of a brand’s experience mind leads to less blame and more forgiveness to a brand. Perception of brand agency mind had no significant effects, and this was expected because the one-dimensional perception of mind (i.e., from no mind to full mind) has essentially emphasized the agency mind of the two-dimensional model (Gray et al., 2007). This is to say that CEOs are generally perceived to have agency and therefore there is minimal variance across the three types of CEOs in term of the brand agency aspect of brand mind. - 94 - Table 12 Study 9: Results of Regression Model of Perceptions of Brand Mind on Moral Judgements CEOs Blame Punishment Forgiveness Perception of Brand Mind .29 ** .38 *** -.34 *** Perception of Brand Mind - Experience -.28 ** -.27 ** .24 ** Perception of Brand Mind - Agency -.01 -.11 .11 Model Evaluation * ** ** *p < .10 **p < .05 ***p < .01 Together the results suggest that when a brand with higher perceived mind is involved in a wrongdoing, consumers are more likely to blame and punish the brand, and less likely to forgive the brand. However, perception of the brand experience aspect of the brand mind may help alter such consequences by leading to less blame and punishment, and more forgiveness. Further, these results show that brand experience mind can be influenced by the leadership perceptions of a CEO endorser. When moving perceived leadership up along inspiring or dominant dimensions, the brand experience aspect of the brand mind will be either increased or reduced correspondingly, resulting in distinct moral judgments from consumers toward brands. This finding provides useful insights for marketing practitioners, namely, it is important to build inspiring leadership that may benefit a brand by positively influencing consumer responses, especially in a brand wrongdoing situation. - 95 - CHAPTER IV CONCLUSION AND GENERAL DISCUSSION This dissertation research investigates how CEOs as brand endorsers influence consumer attitudinal responses differently from conventional brand endorsers such as celebrities and athletes. Accordingly, this research proposes and empirically tests a conceptual model (see Figure 2) that identifies perceived leadership as a distinctive new endorser dimension elicited from a CEO brand endorser (in this work), and predicts that perceived leadership enhances consumers’ perception of brand mind, which in turn plays a critical role in determining consumers’ moral judgments of a brand’s wrongdoing. Boundary conditions including crisis controllability and brand-CEO relationship to the product advertised are discussed. This research contributes to marketing research by developing a new endorser dimension (i.e., perceived leadership) and introducing the concept of mind perception (i.e., brand mind perception) to marketing research. A total of nine studies offers empirical evidence to successfully support the proposals and hypotheses of this research. Studies 1 – 3 successfully develop a perceived leadership scale that is the first to capture how consumers perceive leadership qualities of an organization’s endorser. The approach of this research differs from managerial leadership research, which examines a leader’s behaviors and characteristics observed or experienced directly in a leader-follower dyadic relationship. This set of studies provide a foundational construct, perceived leadership which is then utilized to explore the effects of CEO endorsers in the subsequent studies. - 96 - Two exploratory studies were conducted to gain understanding of perceived leadership (Study 4) and brand mind perception (Study 5). Study 4 reveals that perceived leadership is powerful in predicting how people judge a political figure’s wrongdoing, revealing preliminary evidence of the positive relationship between perceived leadership and blame attribution made by observers. The predictive power in moral judgment differentiates perceived leadership from traditional endorser dimensions such as trustworthiness, expertise, and attractiveness. Study 5 explores the potential of individuals to perceive a brand as having a mind and finds that people have a similar likelihood of perceiving a brand as having a mind, as they do in anthropomorphizing other non-human subjects and identifying their mind. In addition, this likelihood varies depending on individual differences in consumers’ social needs and cognitive styles such as need-for-cognition. This finding confirms that the concept of mind perception is applicable to a brand and can be enhanced by leadership associations. Further, Study 5 shows early evidence that perception of brand mind influences consumers’ moral judgments, suggesting that brand mind perception could be a critical link between perceived leadership and consumer response to brands. Four studies (Studies 6 – 9) provide empirical support to the proposed conceptual model and related hypotheses (Figure 2). Study 6 shows positive effects of a CEO endorser on perceived leadership, resulting in positive consumer attitudes in a marketing communication context; in contrast, Study 7 reveals that perceived leadership from a CEO endorser can lead to more blame and less forgiveness when a brand is involved in wrongdoing. Studies 8 and 9 demonstrate that brand mind perception is a central construct to mediate the effects of perceived leadership (induced here by a CEO - 97 - endorser) on consumers’ moral judgments. Further, CEO-brand relationship (being an endorser CEO from the brand marketed versus being an endorser CEO for a brand with which they have no decision-making affiliation, Study 8) and crisis controllability (Studies 7 and 9) are found as effective moderators. Finally, Study 9 explores how different types of leaders associated with a brand—an inspiring leader or dominant leader—may influence perceptions of brand mind differently. Indeed, the results show that inspiring leadership leads to higher perception of brand experience mind than dominant leadership, resulting in less blame and more forgiving from consumers toward a brand’s wrongdoing. This result shows how the sub-dimensions of perceived leadership - inspiring leadership and dominant leadership may have differential effects on consumer moral judgments. Research Contribution Perceived Leadership and Brand Endorsers. Perceived leadership provides a new perspective (i.e., consumers’ perceptions and consequences) in leadership research, which from the managerial perspective focuses on effective leadership in managing subordinates and enhancing organizational performance. In addition, this dissertation extends brand endorser research by providing a new criterion, perceived leadership, which captures the unique characteristics of CEOs as brand endorsers thus differentiates them from conventional endorsers. Furthermore, this research demonstrates the distinctive power of perceived leadership in contrast to traditional endorser dimensions in predicting brand mind perceptions and consequently consumers’ moral judgments. By doing so, this research - 98 - expands the understanding of endorser effects from traditionally studied message persuasion in marketing communication to consumer moral judgments. Therefore, brand endorser research can be leveraged and thus contribute to knowledge in new marketing circumstances, such as brand transgressions. Viewing endorsers as able to communicate product leadership has significant implications for research. The generalizable contribution is that perceived leadership produces leading brand perceptions and leading brand consequences. While this research only considered the ability of the CEO to communicate leadership, any spokesperson that possesses relevant leadership should function similarly as the CEOs featured in this research. Thus, the endorser must be perceived as having some of the characteristics associated with being a leader and must have a relationship with the brand that communicates the ability to influence the brand’s direction. For example, an athlete, such as basketball player Michael Jordan, that has shown sportsperson leadership, and holds a brand relationship with influence in the naming of the Jordan brand developed by Nike, would likely hold perceived leadership as investigated here. Anthropomorphism and Mind Perception. Most anthropomorphism research in marketing focuses on the product level and is interested in whether a product “looks” or “appears” human. This research contributes to knowledge in brand anthropomorphism by investigating brand mind perception that consumers imbue brands with humanlike minds or mental characteristics. Research on mind perception demonstrates that people intuitively think about other minds, even for nonhuman beings such as a dog or a robot (Gray & Wegner, 2009). To date, however, little research has addressed whether people can perceive a brand as having a mind. Thus, this dissertation contributes to basic - 99 - research on anthropomorphism and mind perception by extending its applications in marketing. Building on mind perception research (i.e. Gray et al., 2007; Morewedge, Preston, & Wegner, 2007; Waytz et al., 2010), this dissertation proposes and defines brand mind perception and its two dimensions, agency and experience. This research finds that the general brand mind is anchored, thus overlaps with agency mind; however, experience mind is unique and influences consumer moral judgments in an opposite direction in contrast with general brand mind. In terms of future research, many questions are opened. Can the experience and agency dimensions of the brand mind be emphasized? Could a long-standing brand communicate its experience? Could a brand’s ability to make changes, for example assembly for disassembly and recycling, be an aspect of the agency mind? Brands as Social Partners. Previous research has explored brand phenomena where brands form social relationships or connections, or engage social interactions with consumers. Fournier and Alvarez (2012) communicate such research as the research in brands as relationship partners. For example, consumers can attribute brands with personalities (Aaker 1997). Consumer-brand congruence plays a critical role in building consumer-brand relationships (Aaker 1999; Bhattacharya and Sen 2003; Malar et al 2011). Consumers relate to brands similarly to how they relate to people (Fournier 1998). They may fall in love with their brands (Batra, Ahuvia, and Bagozzi 2012), become attached to their brands (Thomson, MacInnis, and Park 2005; Park et al. 2010), and develop commitment and loyalty to their brands (Oliver 1999). This stream of research exploring brand social functions, builds from one major assumption that brands can be - 100 - treated as people (Fournier and Alvarez 2012). Perceiving mind is critical for social interactions (Waytz et al. 2010); therefore, this dissertation research in brand mind perception can contribute to research on brand social relationship by enhancing understanding of brand social capabilities. In addition, this research shows evidence that brand mind perception play a critical role in forming consumer moral judgments, thus offering a tool to understand how consumers form moral judgments of brands. Research Limitations and Future Research Opportunities Perceived Leadership and Brand Endorsers. This dissertation identifies a new brand-endorser perception, perceived leadership that reflects a general and stereotypical perception of leadership. This is meaningful because people often rely on heuristic information processing to form attitudes and decisions (Chaiken, 1980), and this perceived leadership can be especially useful in a situation for a new brand or a new CEO of which people have little knowledge. For CEOs who have been active in media and marketing activities and who are well known (e.g., Steve Jobs), it will be interesting to further investigate how perceived leadership interacts with other established characteristics and perceptions of this person. In addition, research should investigate whether and how characteristics of a brand may interact with perceptions of a CEO endorser and together influence consumer perceptions and attitudes. The research is however, limited to the use of CEOs to communicate perceived leadership. Other leaders such as activists or innovators should also be able to influence perceived leadership. Further expanding on broad base of leaders that may be endorsers, Study 1 finds that besides typical leaders such as political figures and company CEOs, a - 101 - broad range of other people, such as athletes and celebrities as well as regular people, were mentioned by some participants as a leader (e.g., Gerard Schwarz - a symphony conductor, LeBron James, Beyoncé, “my Mom”). This indicates that leadership traits are not exclusive to people who are in a titled leadership position. Therefore, research is encouraged to investigate perceived leadership of non-CEO leaders, as well as explore how to enhance perceived leadership in different brand endorsers, thus influencing brand perceptions. For example, will communicating with consumers a story about a celebrity endorser going through childhood struggles and then making headline achievements enhance leadership perception and thus influence consumer perceptions and attitudes? Construal Level and Brand Mind Perception. This dissertation investigates brand endorsers as one factor to influence perception of brand mind. Therefore, research in other brand association factors is highly encouraged. For example, the exploratory study (Study 5) of this research finds that top-of-mind association with brand logo positively impacted the outcome of brand mind perception, but top-of-mind association with purchase experience negatively influenced brand-mind perception. Because “purchase” is a more specific thought of a brand while a “logo” is an abstract representation of a brand, the results seem to suggest that brand mind perception may positively (negatively) correlate with abstract (concrete) brand thoughts. Construal level theory posits that information can appear at a high or low level of abstraction (Trope & Liberman 2003). High-level construal is an abstract representation that contains essential qualities of given information; by contrast, low-level construal is a concrete representation with more details and specific features (Trope & Liberman 2003). Mind is a high-level construal construct since this concept is related to abstract terms - 102 - such as free will, consciousness, intention, and mindfulness (Epley et al. 2007; Gray et al. 2007). Therefore, it is compelling to speculate that when a brand appears in a high-level construal presentation, consumers may be more likely to perceive a brand as having a mind than when a brand appears in a low-level construal presentation. This speculation is not empirically tested within the dissertation. Thus, further research in how construal levels of brand presentations influence perception of brand mind is highly recommended. Limitations of Study Design and Methodology. This research used fictitious stimuli in studies, and only college students or Amazon mTurk participants as the only two sources for data collection, limiting the generalizability of the results. Future research to utilizing field studies or industry data is highly encouraged. In addition, self-reported measures of mind perceptions are adapted to capture perceptions of brand mind in this research. However, the variance of brand mind perceptions, especially experience mind, tends to be small (see Studies 8 and 9). This is possible due to the low tendency of individuals to perceive a brand as having a mind as well as individual differences in this perception (see Study 5). Therefore, future research is encouraged to incorporate individual-difference variables, or new methods to better capture brand mind perception, such as implicit association test (IAT, Greenwald et al. 1998). For example, IAT can be used to ask participants to pair brands with human traits or nonhuman traits (e.g., mental capabilities), and participants are expected to respond faster to pair a brand having higher mind perceptions with mind traits. This method helps capture the effects on participants’ perceptions of brand mind without using explicit questions which may seem “odd” (e.g., to what extend do you agree that this brand can experience emotions?) especially for participants who have higher need for cognition. - 103 - Managerial Implications The findings of this research are useful for marketing practitioners. First, this research responds to the need to understand CEOs as spokespeople for a brand. This marketing phenomenon of more CEOs coming into the spotlight as the face of a brand in marketing communications and generally as communicators about their brands seem to be expanding. This research finds that featuring a CEO as a brand endorser can positively influence consumers’ perception of a brand as a leading brand as well as consumers’ overall attitudes toward a brand and its offerings. Clearly brand managers now have evidence that leadership characteristics of the brand can be supported by the nature of the advertising utilized. While this is not to suggest that leadership perceptions can offset poor quality or lower performance, the research does suggest that communications featuring leaders can tell a brand story in a different way. This knowledge can be leveraged by a new brand or a brand entering a new market to show a brand’s competitive advantage or signal advanced product quality through marketing communications. For brand managers, this research offers new insight that people-brand association can be used to enhance perception of brand mind to increase brand equity. Brand equity reflects the values that a brand itself holds for the products and services that the brand accompanies (Spry et al., 2011). Building a brand as having a mind gives a brand unique characteristics and enables perception of a brand’s social capabilities, thus differentiating such a brand from rival brands. Thus, brand mind perception may be utilized in brand building and consequently have long-term impact on a brand and its business. - 104 - Marketers seeking to anthropomorphize their brand have another possibility. Marketers can focus communications on the mind of the brand. This might be helpful in communicating about the brand’s environmentally friendly composition. The mind of the brand could be “thoughtful” or “determined” in communications. Knowledge of brand mind is also useful for brands facing challenges of public relations. If a brand has a mind, this view implies some accountability for actions and in turn may come with implications for how consumers judge brands. Public interest in corporate impact on society and the environment has grown increasingly. Brands face unexpected challenges from the public’s negative responses (e.g., Starbucks’ Race Together campaign in 2015) or even a public relations crisis (e.g., BP for its massive oil spill in the Gulf of Mexico in 2010). This research lends understanding on how consumers form moral judgments, and thus offers tools brands may utilize in communications and public relations. Specifically, this research shows that enhancing perception of leadership and perception of mind can lead to more blame attribution from consumers toward a brand’s wrongdoing. In contrast, a brand may benefit from using marketing communications to promote inspiring leadership qualities. This research shows that inspiring leadership leads to higher perception of a brand’s experience mind that in turn has positive effects on consumer moral judgments (e.g., less blame and more forgiving). - 105 - APPENDIX A STUDY 4 STIMULI - 106 - APPENDIX B STUDY 6 STIMULI CEO Endorser Athlete Endorser Image courtesy of stockimages at FreeDigitalPhotos.net - 107 - APPENDIX C STUDY 7 STIMULI – NEWS ARTICLE Controllability: Low Controllability: High - 108 - APPENDIX D STUDY 8 STIMULI CEO Endorser / Brand Related Yes CEO Endorser / Brand Related No Employee Endorser / Brand Related Yes Employee Endorser / Brand Related No Image courtesy of stockimages at FreeDigitalPhotos.net - 109 - APPENDIX E STUDY 8 STIMULI – NEWS ARTICLE - 110 - APPENDIX F STUDY 9 STIMULI Employee – Low Leadership CEO – High Leadership CEO – Inspiring Leadership CEO – Dominant Leadership Image courtesy of stockimages at FreeDigitalPhotos.net - 111 - APPENDIX G STUDY 9 STIMULI - NEWS ARTICLE Controllability: High Controllability: Low - 112 - APPENDIX H MEASURES SUMMARY Study 3 Traditional Endorser Dimensions (7-point bipolar scale; Ohanian 1990) - Trustworthiness: Undependable – Dependable Dishonest – Honest Unreliable – Reliable Insincere – Sincere Untrustworthy – Trustworthy - Expertise Not an expert – Expert Inexperienced – Experienced Unknowledgeable – Knowledgeable Unqualified – Qualified Unskilled – Skilled - Attractiveness Unattractive – Attractive Not Classy – Classy Ugly – Beautiful Plain – Elegant Not sexy - Sexy New Perceived Leadership Scale (7-point bipolar scale) - Inspiring Not inspiring – Inspiring Ineffective - Effective Irresponsible – Responsible Uncourageous – Courageous - Dominant Non-dominant – Dominant Powerless – Powerful Non-authoritative – Authoritative General Perception as a Leader (endpoints: 1 = strongly disagree; 7 = strongly agree) - “Please indicate to what extent you agree with the following statements about the person you just evaluated: 1. The person is a leader; 2. This person typifies what a leader is; 3. This person has leadership qualities.” - 113 - Study 4 Single Item Moral Judgment Measure: Blame, Disappointment, Forgiveness, and Intention - “Please indicate to what extent you agree with the statements (endpoints: 1 = strongly disagree; 7 = strongly agree) 1. I would blame this person if he or she did something wrong (blame); 2. If this person did something wrong, I would be very disappointed (disappointment) 3. I would forgive this person if he or she apologized for a wrongdoing (forgiveness).” - Intention (endpoints: 1 = extremely unlikely; 7 = extremely likely) 4. “If this person did something wrong, how likely would you believe he or she did it intentionally?” General Attitudes (on a seven-point bipolar scale; Mitchell & Olson, 1981) - “Please indicate your general attitudes toward this person. Unfavorable – Favorable Bad – Good Dislike – Like” Study 5 Location - “Please indicate your location in the U.S.A. 1. West 2. South 3. Northeast 4. Midwest 5. Hawaii 6. Alaska 7. Other territories of the United States (Guam, Northern Mariana Islands, Puerto Rico, United States Virgin Islands 8. Not from U.S.A” Religion - What best describes your religious affiliation (or non-affiliation belief)? 1. Christian - Protestant (includes Baptist, Methodist, Lutheran, Presbyterian, Pentecostal, Non-denominational, etc.) 2. Christian - Catholic 3. Jewish 4. Muslim 5. Buddhist - 114 - 6. Hindu 7. Spiritual but not religious 8. Agnostic (not sure) 9. Atheist (believe there is no god) 10. Other ____________________ Race - “Please indicate which of the following best represents your racial or ethnic heritage? Choose all that apply: 1. Non-Hispanic White or Euro-American 2. Black, Afro-Caribbean, or African American 3. Latino or Hispanic American 4. East Asian or Asian American 5. South Asian or Indian American 6. Middle Eastern or Arab American 7. Native American or Alaskan Native 8. Hawaiian / Pacific Islander 9. Other” Individual Differences (Endpoints:1= strongly disagree; 5= strongly agree) 1) Self-construal Inter-dependence Scale (Escalas & Bettman,2005) - “Please indicate to what extent you agree with the following statements. 1. It is important for me to respect decisions made by the group I belong. 2. I often have the feeling that my relationships with others are more important than my own accomplishments.” 2) Loneliness (Gierveld & Tilburg, 2006) - “The following statements may or may not apply to your situation, the way you feel now. Please indicate to what extent do you agree or disagree with the following statements on whether they can apply to your situation, the way you feel now. 1. There are plenty of people I can rely on when I have problems. 2. There are many people I can trust completely. 3. There are enough people I feel close to. 4. I experience a general sense of emptiness. (reverse coded) 5. I miss having people around. (reverse coded) 6. I often feel rejected. (reverse coded)” - 115 - 3) Need-for-cognition (Wood and Swait, 2002) - “The following statements may or may not describe how you think. There are no right or wrong answers. 1. I only think as hard as I have to. 2. The notion of thinking abstractly is not appealing to me. 3. The idea of relying on thought to get my way to the top does not appeal to me. 4. I would rather do something that requires little thought than something that is sure to challenge my thinking abilities. 5. I try to anticipate and avoid situations where there is a likely chance I'll have to think in depth about something.” Study 6 Attitude toward Ads (on a seven-point bipolar scale; Mitchell & Olson, 1981) - “Please indicate your general attitudes toward the advertisement. 1. Unfavorable – Favorable 2. Bad – Good 3. Dislike – Like” Leading Brand Perception (endpoints: 1 = strongly disagree; 7 = strongly agree) - “Please indicate to what extent you agree with the following statements about the person you just evaluated: 1. Hudson is a leader in its industry. 2. Hudson is a leading brand. 3. Hudson shows leadership.” Study 7 Blame (endpoints: 1 = strongly disagree; 7 = strongly agree; De Matos, & Rossi, 2007) - “Please indicate to what extent you agree with the following statements. 1. I will blame Hudson the brand for the poisonous materials used in its products. 2. I believe Hudson the brand is responsible for the poisonous materials in its products. 3. It is Hudson the brand’s fault for using the poisonous materials in its products” - 116 - Punishment (endpoints: 1 = not at all; 7 = very much) - “How much do you think the brand deserves punishment?” Forgiveness (endpoints: 1 = not at all; 7 = very much) - “How likely would you forgive the brand if the brand apologized?” Crisis Controllability (endpoints: 1 = strongly disagree; 7 = strongly agree; adapted from McAuley, Duncan, & Russell, 1992) - “Please indicate to what extent you agree with the following statements. 1. The causes of poisonous materials were controlled by the company; 2. The causes of poisonous materials were always present; 3. The product crisis happened in a process which was under control of the company.” Study 8 Brand Mind Perception (adapted from Morewedge, Preston, & Wegner, 2007) - “Please indicate the degree to which Janssen appears to have a mind of its own” (1 = definitely has no mind; 7 = definitely has a mind); - “Please indicate to what extent you agree or disagree with the statement that Janssen has mental capabilities (e.g., thinking, making plans, communication)” (1 = strongly disagree; 7 = strongly agree). Study 9 Brand Agency Mind (endpoints: 1 = not at all; 7 = very much; adapted from Gray et al., 2007) - “To what degree can Janssen make plans and work toward goals? - To what degree is Janssen capable of thinking? - To what degree is Janssen capable of making decisions?” Brand Experience Mind (endpoints: 1 = not at all; 7 = very much; adapted from Gray et al., 2007) - “To what degree is Janssen capable of having experiences and being aware of things?” - To what degree can Janssen feel joy? - To what degree can Janssen experience embarrassment?” - 117 - REFERENCES CITED Aaker, J. (1997). Dimensions of brand personality. Journal of Marketing Research, 34(3), 347-356. Aaker, J. (1999). The malleable self: The role of self-expression in persuasion. Journal of Marketing Research, 36(1), 45-57. Aaker, J., Fournier, S., & Brasel, S. A. (2004). When good brands do bad. Journal of Consumer Research, 31(1), 1-16. Aggarwal, P., & McGill, A. (2007). Is that car smiling at me? Schema congruity as a basis for evaluating anthropomorphized products. Journal of Consumer Research, 34(4), 468–479. Aggarwal, P & McGill, A.L. (2012), When brands seem human, do humans act like brands? Automatic behavioral priming effects of brand anthropomorphism, Journal of Consumer Research, 39(2), 307-23. Albert, N., Merunka, D., & Valette-Florence, P. (2008). When consumers love their brands: Exploring the concept and its dimensions. Journal of Business Research, 61(10), 1062–1075. Arbuckle, J. L. (2013). IBM SPSS Amos 22 user’s guide. Crawfordville, FL: Amos Development Corporation. Avolio, B. J., Bass, B. M., & Jung, D. I. (1999). Re‐examining the components of transformational and transactional leadership using the multifactor leadership. Journal of Occupational and Organizational Psychology, 72(4), 441- 462. Ball, D. W. (1975). A note on method in the sociological study of sport. Sport and Social Order: Contributions to the Sociology of Sport. Reading, MA: Addison-Wesley. Baron-Cohen, S., Leslie, A. M., & Frith, U. (1985). Does the autistic child have a “theory of mind”? Cognition, 21(1), 37-46. Bass, B. M. (1985). Leadership and Performance beyond Expectations. New York, NY: Collier Macmillan. Bass, B. M. (1990). From transactional to transformational leadership: Learning to share the vision. Organizational Dynamics, 18(3), 19-31. Bass, B. M., & Avolio, B. J. (1994). Improving Organizational Effectiveness Through Transformational Leadership. Thousand Oaks, CA: Sage. - 118 - Batra, R., Ahuvia, A., & Bagozzi, R. P. (2012). Brand love. Journal of Marketing, 76(2), 1–16. Bhattacharya, C. B., & Sen, S. (2003). Consumer-company identification: A framework for understanding consumers’ relationships with companies. Journal of Marketing, 67(2), 76-88. Brewer, M. B., & Gardner, W. (1996). Who is this "We"? Levels of collective identity and self representations. Journal of Personality and Social Psychology, 71(1), 83- 93. Burns, J. M. (1978). Leadership. New York, NY: Harper & Row Cain, M. D., & McKeon, S. B. (2016). CEO personal risk-taking and corporate policies. Journal of Financial and Quantitative Analysis, 51(01), 139-164. Carlson, B. D., & Donavan, D. T. (2008). Concerning the effect of athlete endorsements on brand and team-related intentions. Sport Marketing Quarterly, 17(3), 154-162. Carlyle, T. (1849). On Heroes, Hero-worship, and the Heroic in History. Boston, MA: Houghton-Mifflin. Chaiken, S. (1980). Heuristic versus systematic information processing and the use of source versus message cues in persuasion. Journal of Personality and Social Psychology, 39(5), 752-766. Chandler, J., & Schwarz, N. (2010). Use does not wear ragged the fabric of friendship: Thinking of objects as alive makes people less willing to replace them. Journal of Consumer Psychology, 20(2), 138–145. Chelladurai, P., & Saleh, S. D. (1980). Dimensions of leader behavior in sports: Development of a leadership scale. Journal of Sport Psychology, 2(1), 34-45. Chien, P. M., Cornwell, T. B., & Pappu, R. (2011). Sponsorship portfolio as a brand- image creation strategy. Journal of Business Research, 64(2), 142-149. Childers, T. L. (1986). Assessment of the psychometric properties of an opinion leadership scale. Journal of Marketing Research, 23(2), 184-188. Churchill Jr, G. A. (1979). A paradigm for developing better measures of marketing constructs. Journal of Marketing Research, 16(1), 64-73. Cohen, A.R., Stotland, E., & Wolfe, D.M., (1955). An experimental investigation of need for cognition. Journal of Abnormal and Social Psychology, 51(2), 291–294 - 119 - Collins, A. M., & Loftus, E. F. (1975). A spreading-activation theory of semantic processing. Psychological Review, 82(6), 407-428. Cornwell, T. B., Roy, D. P., & Steinard, E. A. (2001). Exploring managers' perceptions of the impact of sponsorship on brand equity. Journal of Advertising, 30(2), 41- 51. Coombs, W. T., & Holladay, S. J. (Eds.). (2011). The Handbook of Crisis Communication (Vol. 22). John Wiley & Sons. Dawar, N., & Pillutla, M. M. (2000). Impact of product-harm crises on brand equity: The moderating role of consumer expectations. Journal of Marketing Research, 37(2), 215-226. Edelman, G. M. (1992). Bright Air, Brilliant Fire: On the Matter of the Mind. New York, NY: Basic books. Epley, N., Schroeder, J., & Waytz, A. (2013). Motivated mind perception: Treating pets as people and people as animals. Objectification and (De) Humanization, 127-152. Epley, N., Waytz, A., & Cacioppo, J. T. (2007). On seeing human: A three-factor theory of anthropomorphism. Psychological Review, 114(4), 864–86. Erdogan, B. Z. (1999). Celebrity endorsement: A literature review. Journal of Marketing Management, 15(4), 291-314. Escalas, J. E., & Bettman, J. R. (2005). Self‐construal, reference groups, and brand meaning. Journal of Consumer Research, 32(3), 378-389. Field, A. (2013). Discovering statistics using IBM SPSS statistics. Thousand Oaks, CA: Sage. Flynn, L. R., Goldsmith, R. E., & Eastman, J. K. (1994). The king and summers opinion leadership scale: Revision and refinement. Journal of Business Research, 31(1), 55-64. Fornell, C., & Larcker, D. F. (1981). Structural equation models with unobservable variables and measurement error: Algebra and statistics. Journal of Marketing Research, 18(3), 382-388. Fournier, S. (1998). Consumers and their brands: Developing relationship theory in consumer research. Journal of Consumer Research, 24(4), 343–353. Fournier, S., & Alvarez, C. (2012). Brands as relationship partners: Warmth, competence, and in-between. Journal of Consumer Psychology, 22(2), 177-185. - 120 - Fujita, K., Henderson, M. D., Eng, J., Trope, Y., & Liberman, N. (2006). Spatial distance and mental construal of social events. Psychological Science, 17(4), 278-282. Gierveld, J. D. J., & Tilburg, T. V. (2006). A 6-item scale for overall, emotional, and social loneliness: Confirmatory tests on survey data. Research on Aging, 28(5), 582-598. Gray, H. M., Gray, K., & Wegner, D. M. (2007). Dimensions of mind perception. Science, 315(5812), 619. Gray, K., Young, L., & Waytz, A. (2012). Mind perception is the essence of morality. Psychological Inquiry, 23(2), 101-124. Gray, K., & Wegner, D. M. (2009). Moral typecasting: Divergent perceptions of moral agents and moral patients. Journal of Personality and Social Psychology, 96(3), 505–20. Gray, K., & Wegner, D. M. (2011). To escape blame, don’t be a hero - be a victim. Journal of Experimental Social Psychology, 47(2), 516–519. Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. (1998). Measuring individual differences in implicit cognition: the implicit association test. Journal of Personality and Social Psychology, 74(6), 1464-1480. Haleblian, J., & Finkelstein, S. (1993). Top management team size, CEO dominance, and firm performance: The moderating roles of environmental turbulence and discretion. Academy of Management Journal, 36(4): 844-863. Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: the Guilford Press. Herr. P.M., (1989) Priming price: Prior knowledge and context effects. Journal of Consumer Research, 16(1): 67-75. House, R. J., & Aditya, R. N. (1997). The social scientific study of leadership: Quo vadis?. Journal of Management, 23(3), 409-473. Hovland, C. I., Janis I. K., and Kelley, H.H. (1953). Communication and persuasion, New Haven, CT: Yale University Press. Judge, T. A., Bono, J. E., Ilies, R., & Gerhardt, M. W. (2002). Personality and leadership: A qualitative and quantitative review. Journal of Applied Psychology, 87(4), 765- 780. - 121 - Judge, T. A., Piccolo, R. F., & Kosalka, T. (2009). The bright and dark sides of leader traits: A review and theoretical extension of the leader trait paradigm. The Leadership Quarterly, 20(6), 855-875. Keller, K. L. (1993). Conceptualizing, measuring, and managing customer-based brand equity. Journal of Marketing, 57(1), 1-22. Kim, S., & McGill, A. L. (2011). Gaming with Mr. Slot or gaming the slot machine? Power, anthropomorphism, and risk perception. Journal of Consumer Research, 38(1), 94–107. King, C. W., & Summers, J. O. (1970). Overlap of opinion leadership across consumer product categories. Journal of Marketing Research, 7(1), 43-50. Kirkpatrick, S. A., & Locke, E. A. (1996). Direct and indirect effects of three core charismatic leadership components on performance and attitudes. Journal of Applied Psychology, 81(1), 36-51. Klein, J., & Dawar, N. (2004). Corporate social responsibility and consumers' attributions and brand evaluations in a product–harm crisis. International Journal of Research in Marketing, 21(3), 203-217. Landwehr, J. R., McGill, A. L., & Herrmann, A. (2011). It’s got the look: The effect of friendly and aggressive “facial” expressions on product liking and sales, Journal of Marketing, 75(3), 132–146. Liu, B. F., Austin, L., & Jin, Y. (2011). How publics respond to crisis communication strategies: The interplay of information form and source. Public Relations Review, 37(4), 345-353. MacKenzie, S. B., Podsakoff, P. M., & Rich, G. A. (2001). Transformational and transactional leadership and salesperson performance. Journal of the Academy of Marketing Science, 29(2), 115-134. Malär, L., Krohmer, H., Hoyer, W. D., & Nyffenegger, B. (2011). Emotional brand attachment and brand personality: The relative importance of the actual and the ideal self. Journal of Marketing, 75(4), 35-52. Markus, H. & Shinobu, K. (1991). Culture and the self: Implications for cognition, emotion, and motivation, Psychological Review, 98(2), 224–53. McAuley, E., Duncan, T. E., & Russell, D. W. (1992). Measuring causal attributions: The revised causal dimension scale (CDSII). Personality and Social Psychology Bulletin, 18(5), 566-573. - 122 - McCarthy, M. (2013). Tesla generates small sales, big buzz without paid ads. ADAGE, Retrieved from http://adage.com/article/news/tesla-generates-small-sales-big- buzz-paid-ads/241994/ McGuire, W. J. (1964). Some contemporary approaches. Advances in Experimental Social Psychology, 1, 191-229. McGuire, W. J. (1985). Attitudes and attitude change. Handbook of Social Psychology (Gardner, L., & Aronson, E. eds. Vol. 2, pp. 233-346). New York, NY: Random House. Morewedge, C. K., Preston, J., & Wegner, D. M. (2007). Timescale bias in the attribution of mind. Journal of Personality and Social Psychology, 93(1), 1-11. Morhart, F. M., Herzog, W., & Tomczak, T. (2009). Brand-specific leadership: Turning employees into brand champions. Journal of Marketing, 73(5), 122-142. Mowen, J. C. (1990). Consumer Behavior. New York, NY: Macmillian Myers, J. H., & Robertson, T. S. (1972). Dimensions of opinion leadership. Journal of Marketing Research, 9(1), 41-46. Nunnally, J. (1978). Psychometric Methods. New York, NY: McGraw-Hill O’Connell, V., & O’Sullivan, D. (2011). The impact of customer satisfaction on CEO bonuses. Journal of the Academy of Marketing Science, 39(6), 828-845. Ohanian, R. (1990). Construction and validation of a scale to measure celebrity endorsers' perceived expertise, trustworthiness, and attractiveness. Journal of Advertising, 19(3), 39-52. Ohanian, R. (1991). The impact of celebrity spokespersons' perceived image on consumers' intention to purchase. Journal of Advertising Research. 31(1), 46-54. Oliver, R. L. (1999). Whence consumer loyalty? Journal of Marketing, 63(4), 33-44. Oxenfeldt, A.R. and Swan, C. (1964), Management of the Advertising Function, Belmont, CA: Wadsworth. Park, W. C., MacInnis, D. J., Priester, J., Eisingerich, A. B., & Iacobucci, D. (2010). Brand attachment and brand attitude strength: Conceptual and empirical differentiation of two critical brand equity drivers. Journal of Marketing, 74(6), 1- 17. Park, C. W., Jaworski, B. J., & Maclnnis, D. J. (1986). Strategic brand concept-image management. Journal of Marketing, 50(4), 135-145. - 123 - Parrino, R. (1997). CEO turnover and outside succession a cross-sectional analysis. Journal of Financial Economics, 46(2), 165-197. Peterson, R. S., Smith, D. B., Martorana, P. V., & Owens, P. D. (2003). The impact of chief executive officer personality on top management team dynamics: One mechanism by which leadership affects organizational performance. Journal of Applied Psychology, 88(5), 795-808. Pham, S. (2017). Warren Buffett is the face of Cherry Coke in China. CNN Money, Retrieved from http://money.cnn.com/2017/04/04/news/companies/warren- buffett-cherry-coke-china/ Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research Methods, 36(4), 717-731. Premack, D., & Woodruff, G. (1978). Does the chimpanzee have a theory of mind? Behavioral and Brain Sciences, 1(4), 515-526. Quirk, M. P., and Fandt, P. M. (2000) The 2nd Language of Leadership. Mahwah, NJ: Lawrence Erlbaum Associates. Rogers, E. M., & Cartano, D. G. (1962). Methods of measuring opinion leadership. Public Opinion Quarterly, 26(3), 435-441. Rubin, R. S., Munz, D. C., & Bommer, W. H. (2005). Leading from within: The effects of emotion recognition and personality on transformational leadership behavior. Academy of Management Journal, 48(5), 845-858. Saucier, G. (1994). Mini-Markers: A brief version of Goldberg's unipolar Big-Five markers. Journal of Personality Assessment, 63(3), 506-516. Schriesheim, C. A., & Stogdill, R. M. (1975). Differences in factor structure across three versions of the Ohio State leadership scales. Personnel Psychology, 28(2), 189- 206. Shocker, A. D., & Srinivasan, V. (1979). Multiattribute approaches for product concept evaluation and generation: A critical review. Journal of Marketing Research, 16(2), 159-180. Singelis, T. M. (1994). The measurement of independent and interdependent self- construals. Personality and Social Psychology Bulletin, 20(5), 580-591. Sozzi, B. (2016) Why Doug McMillon is the first-ever Walmart CEO to star in a TV commercial. The Street, Retrieved from https://www.thestreet.com/story/13682977/1/why-doug-mcmillon-is-the-first- ever-walmart-ceo-to-star-in-a-tv-commercial.html - 124 - Spry, A., Pappu, R., & Cornwell, T. B. (2011). Celebrity endorsement, brand credibility and brand equity. European Journal of Marketing, 45(6), 882-909. Stogdill, R. M., & Coons, A. E. (1957). Leader Behavior: Its Description and Measurement. Bureau of Business Research, College of Commerce and Administration, Ohio State University. Strohminger, N., & Nichols, S. (2014). The essential moral self. Cognition, 131(1), 159- 171. Swan, C. J. (1964). Management of the Advertising Function. Boston, MA: Wadsworth Publishing Company. Thomson, M., MacInnis, D. J., & Park, C. W. (2005). The ties that bind: Measuring the strength of consumers' emotional attachments to brands. Journal of Consumer Psychology, 15(1), 77–91. Trope, Y., & Liberman, N. (2003). Temporal construal. Psychological Review, 110(3), 403–421. Wayne, S. J., Shore, L. M., & Liden, R.C. (1997) Perceived organizational support and leader-member exchange: A social exchange perspective. Academy of Management Journal, 40(1), 82-111. Waytz, A., Gray, K., Epley, N., & Wegner, D. M. (2010). Causes and consequences of mind perception. Trends in Cognitive Sciences, 14(8), 383-388. Wood, S. L., & Swait, J. (2002). Psychological indicators of innovation adoption: Cross- classification based on need for cognition and need for change. Journal of Consumer Psychology, 12(1), 1-13. Yermack, D. (1995). Do corporations award CEO stock options effectively? Journal of Financial Economics, 39(2), 237-269. Yi, Y. (1990). A critical review of consumer satisfaction. Review of Marketing, 4(1), 68- 123. Zook, C., & Allen, J. (2016). The Founder's Mentality: How to Overcome the Predictable Crises of Growth. Brington, MA: Harvard Business Review Press. Zaccaro, S. J., Kemp, C., & Bader, P. (2004). Leader traits and attributes. The Nature of Leadership (Day, D. V., & Antonakis, J. eds. pp. 101-124). Thousand Oaks, CA: Sage