2013 Xi Jiang 2

Transcription

1 A CROSS-CULTURAL STUDY OF CRISIS RESPONSE STRATEGY EFFECTIVENESS By XI JIANG A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS IN MASS COMMUNICATION UNIVERSITY OF FLORIDA

2 2013 Xi Jiang 2

3 For my Mom and Dad, whose love truly enlightened my life. 3

4 ACKNOWLEDGMENTS I would like to acknowledge the advice and guidance of Dr. Sora Kim, committee chair, without whose help this study would not have been completed. I also thank the members of my graduate committee, Dr. Juan-Carlos Molleda and Dr. Mary Ann Ferguson, for all their advice, encouragement and knowledge. 4

5 TABLE OF CONTENTS ACKNOWLEDGMENTS... 4 LIST OF TABLES... 7 LIST OF FIGURES... 8 ABSTRACT... 8 CHAPTER 1 INTRODUCTION Crisis Communication across Culture Culture Identity, Negative Emotion & Attributions of Blame Effectiveness of Crisis Response Strategy in Terms of Emotion and Attribution of Blame: A Public-based Approach Research Gap Purpose of Study Importance of Study LITERATURE REVIEW Emotion: A Public-based Perspective of Crisis Communication How Does Organizational Crisis Affect the Publics Emotion: Understanding the Influencing Process Stressful Situations Elicit Negative Emotion Emotion and Crisis Responsibility Integrated Crisis Mapping (ICM) Model Attribution of Blame Situational Crisis Communication Theory Cultural Issues in Crisis Communication Cultural Issues, Blame of Attribution, and Negative Emotions Hypotheses and Research Questions METHODOLOGY Participants Procedure Stimulus Material Crisis Type Crisis Response Strategy Measures Power Distance Scale Negative Emotions RESULTS

6 Manipulation Check Test of Hypotheses DISCUSSION IMPLICATION AND LIMITATION APPENDIX A STIMULI B CHINESE VERSION STIMULI C APPENDIX C LIST OF REFERENCES BIOGRAPHICAL SKETCH

7 Table LIST OF TABLES page 2-1 Crisis response strategies by postures Mean and standard deviation of blame attribution in China and the US Means and standard deviation of negative emotions for crisis response strategies in the United States and China

8 Figure LIST OF FIGURES page 2-1 Revised Integrated Crisis Mapping model Contrast China s and America s five cultural dimensions Interaction between nationality and corporate response in reducing publics attribution of blame

9 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts in Mass Communication A CROSS-CULTURAL STUDY OF CRISIS RESPONSE STRATEGY EFFECTIVENESS Chair: Sora Kim Major: Mass Communication By Xi Jiang August 2013 This study employed a 2 (Chinese and the US) 4 (apology, justification, denial and control) experiment to explore the crisis response strategies effectiveness in different cultures in terms of mitigating publics negative emotion and attribution of blame. The results showed that the Chinese respondents in this experiment displayed a higher level of negative emotion than the US respondents. Also, different corporate crisis response strategies elicited participants different level of blame attribution. The blame attribution also varies with culture. This study contributes to the practice of optimize crisis response strategy in different cultures. 9

10 CHAPTER 1 INTRODUCTION The word crisis may be among the most vexing words for public relations practitioners. Certainly, strategic crisis communication is one of the biggest challenges in the global marketplace (Taylor, 2000). As multinational corporations expand across the globe and each nation has specific cultural dynamic, public relations practitioners must consider cultural variances when delivering messages (Wertz & Kim, 2010). Crisis Communication across Culture Public relations scholars have found that public relations is practiced in substantially different ways in different countries, and often reflecting cultural differences (Grunig, Grunig, & Dozier, 2006). Cultural differences also affect the work of crisis management. However, crisis communication researchers paid little attention to how cultural issues affected crisis communication until several multinational corporations failed to use culturally appropriate crisis response strategies in several prominent instances, for instance, the contaminated Coca-Cola crisis in Europe (see Taylor, 2000). After some pioneering works addressing the importance of cultural issues in crisis management, for example Taylor s paper about the Coca-Cola crisis and Haruta and Kirk s (2003) paper about cultural issues in airline crisis, more scholars d that managers of efficient cross-cultural communication conflicts are more likely to show respect to cultural differences. For example, although many American scholars implied that a company adopting a no-comment crisis communication strategy might be perceived negatively in the U.S. (Davis & Holtgraves, 1980), remaining silent is often seen as an act of wisdom in Chinese culture and therefore Lee (2004) pointed out in her experimental study that Chinese respondents showed more tolerance towards no-comment. 10

11 Existing literature indicates that culture is, indeed, a vital variable that influences not only the response of key publics, but also organizational strategy and communication effectiveness in crisis communication (Taylor, 2000; Haruta & Kirk, 2003; Giebels & Taylor, 2009; Wertz & Kim, 2010). Intercultural communication theories are widely used as the critical frameworks to substantiate and explain cultural differences in crisis communication. For instance, Taylor (2000) and Wertz and Kim (2010) adopted Hofstede s cultural dimensions to demonstrate the link between culture, the way an organization deals with crisis communication, and how the public responds to that communication. According to Taylor (2000), a high uncertainty avoidance index, which is one of Hofstede s (2012) six cultural dimensions, index is positively related to publics low risk taking, eagerness to know the result, and relatively strong reaction. Another cultural dimension, power distance, is positively related to citizens low trust in authority, and such distrust is more likely to ignite eventual conflict between a well-known organization and the public (Taylor, 2000). Confucian dynamism, known as long-term orientation (Hoffstede, 1991), on the other hand, is positively related to the organization whose crisis response includes more full apology messages (Wertz & Kim, 2010). Utilizing Hall s high and low context culture theory, Giebel & Taylor (2009) examined negotiators responses to persuasive argument in crisis negotiations over time. They revealed that, compared to high-context perpetrators, low-context perpetrators were found to use more persuasive arguments and to respond to persuasive arguments in a compromising way. Based on those findings, Huang and Bedford (2009) wrote that understanding cross-cultural differences in conflict management style and crisis communicative strategy is essential to the practice of public diplomacy and public 11

12 relations, especially given the globalized business economy and the proliferation of international trade ments (p. 565). Culture Identity, Negative Emotion & Attributions of Blame Generally speaking, Easterners and Westerners share a similar cognitive process, which implies that a universal psychological state would lead them to arriving at the same judgment of communication during a crisis (Gardner, 1985). However, crosscultural psychology works suggest that differences between Western and Eastern cognitive pattern exist with regard to emotion and attributions of blame (An et al., 2010; Anagondahalli & Turner, 2012). During a crisis, key publics will automatically look for causes of the threat, and attribute blame to the responsible organization (Coombs, 2005). It is argued that a higher level of collectivism is positively related to a higher level of public attribution of blame of an organization (An et al., 2010; Menon, et al., 1999). For example, Menon et al. (1999) found out that East Asians are more likely to attribute outcomes of accidents to a group rather than to an individual. One possible explanation of different levels of attribution is that Asians view an organization as an extended family, and it's not appropriate to blame a member of a family therefore it is not right to blame the members of an organization (An et al., 2010). In contrast to Asians, Americans have a different pattern of blame. They tend to blame an organization right away, then transition to blaming certain individuals (Zemba, Young, & Morris, 2006). Another point that is worth noting is that stakeholders with higher level of collectivism tend to show more negative emotion towards an organization than those from individualistic culture (An et al., 2010). For example, Anagondahalli & Turner 12

13 (2012) reported that Asians respondents experienced more anxiety than American respondents in an organizational crisis, and An et al. (2010) revealed that Koreans showed more anger than their American counterparts in the same crisis situation. Effectiveness of Crisis Response Strategy in Terms of Emotion and Attribution of Blame: a Public-based Approach Previous crisis literature tends to view crisis communication strategies from an organization-focused perspective; the dominant theories in this field are Coombs Situational Crisis Communication Theory (SCCT) and Benoits Image Repair Theory (Benoit, 1997; Coombs & Holladay, 1996; Coombs & Holladay, 2004), and these two theories are built around protecting the organization s reputation during a crisis (Coombs, 2006). However, Coombs (2006) pointed out that the organization-based perspective of crisis response strategies overlooked the stakeholders perception and argued for the need to view organizational crisis strategies from the stakeholder s perspective. Currently, two approaches are offered to examine crisis strategy from the audience's standpoint: the public emotion-based perspective and public attribution of blame (Jin, 2010; Jin, 2009; Jin, Pang, & Cameron, 2007; Anagondahalli & Turner, 2012). These two perspectives are intertwined, and one can affect the other (Jin, 2009). The first perspective, public emotion-based perspective, has evolved into a hot topic in the academic field in recent years. Until now, several studies testified to emotion s importance in mass communication. For instance, scholars have looked into consumers emotional responses to advertising (Derbaix, 1996), employees' preference of a leader s emotional expressiveness (Madera & Smith, 2009), and the role of emotion in computer-mediated communication (Derks, Fischer, & Bos, 2008). In order to better 13

14 understand the public's mind and to shape appropriate strategy, crisis communication experts integrated the emotion-related approach. Both professional public-relations strategists and scholars believe that managing emotions is a missing step in crisis communication (Loomis, 2008; Yeomans, 2007). A school of crisis communication scholars asserts that although organization-based crisis communication (for instance, Coombs Situational Crisis Communication Theory) is of great significance in crisis communication and emotion-focused crisis communication is more generic and systemic, understanding the diverse and varied emotions that are likely to be experienced by the stakeholders in a crisis is helpful for public practitioners to create the best response (Jin, Pang, & Cameron, 2012). Emotion-driven strategy is meaningful for public relations practitioners and researchers to understand what emotional upheavals might occur to the public, so that certain strategies can be adopted aiming at their specific needs (Jin, Pang, & Cameron, 2007, p. 83). In addition, emotion is highly related to people s evaluation of an organization and its communication. Since an initial emotional reaction dominates people s affective evaluation in a crisis (Choi & Lin, 2009), negative emotions elicited at that time are very likely to weaken an organization s already existing positive image (Coombs & Holladay, 2006). Compared to emotion, the second perspective, attribution of blame, has received more attention in past years (Bradford & Garrett, 1995; Coombs & Holladay, 1996). Because the two key characteristics of crises that they are unexpected and negative (Coombs, 2006) are also the key traits that drive people s need to search for causes of an event in Attribution Theory (Weiner, 1985), it is reasonable to connect crisis with 14

15 Attribution Theory. In Attribution Theory, the threat of a crisis largely comes from the public's blaming of the organization that is responsible for the crisis (Coombs, 2006). For instance, crisis experts found that respondents attribute more blame to an organization perceived as responsible for the crisis, and would be less likely to have a good impression about this organization (Jin, 2009). Based upon the public's attribution of blame, managers can better assess the threat (Coombs, 1995). Research Gap While most studies were to guide practitioners in choosing a crisis response strategy from an organizational point of view (i.e., trying to find a way to reduce financial loss or legal liability), a research gap exists regarding selecting organizational crisis strategy from a public-based standpoint. This gap needs to be closed so that an organization can match crisis response strategy with the psychological needs and impressions of the public and create a successful crisis communication campaign (Coombs, 2006; Jin, 2010). Additionally, intercultural communication studies have served as beacons in terms of identifying emotional universality and nuanced differences between cultures (Andersen, 2009). Clearly, cultural preferences, local traditions, demographic patterns, and cultural dynamics all affect the effectiveness of crisis response strategy (Taylor, 2000; Haruta & Kirk, 2003; Giebels & Taylor, 2009; Wertz & Kim, 2010). Purpose of Study In order to create a useful rubric for evaluation of crisis communication strategy effectiveness, this paper tests and compares the effectiveness of crisis response strategies in different cultural settings. First, this study explores which crisis response 15

16 strategy is more effective in mitigating negative public emotion, and if there is any cultural difference between Chinese and American audiences response in terms of such effectiveness. Second, this study investigates which strategy would be more effective in reducing the publics attributions of blame, and if there is any cultural difference between Chinese and American audiences response in terms of such effectiveness. Finally, this study examines if there is any relationship between Hofstede s cultural dimension of power distance and public responses such as negative emotion and blame attribution. Importance of Study Given that alleviating the public's negative emotion and blame attribution is essential for restoring an organizational image (Coombs & Holladay, 2004), the effectiveness of a certain strategy deserves more academic attention in public relations. This study focuses on testing the negative emotions and levels of blame of key publics. Unlike a negative attitude, which may occurs when a crisis is not handled appropriately, some negative emotions, such as anxiety and alertness, are identified as default emotions in crisis and appear automatically in stressful situations (Choi & Lin, 2009; Jin, Pang, & Cameron, 2012). This means that no matter what post-crisis strategy is adopted, the publics first reaction in a crisis scenario could be always to show negative emotion. This study provides practitioners with a public-oriented perspective to estimate the effectiveness of a crisis response strategy, and aims to offer practitioners an optimal choice when choosing a strategy. Also, by providing scholars comparative results from an experiment, this research gives crisis managers an empirical and insightful system to cope with intercultural crisis. 16

17 Simultaneously, the findings can help multinational corporations management of crosscultural crisis communication. 17

18 CHAPTER 2 LITERATURE REVIEW Emotion: A public-based perspective of crisis communication Crisis communication is the collection and dissemination of information by the crisis management team (Coombs, 2005, p. 221). In recent decades, the crisis communication field has grown fast and produced many volumes of research findings. Researchers mainly conducted case studies in the late 1990s, when they conducted many content analysis studies based upon real-world cases (Kim & Cameron, 2011). Since then, theorists have been building a theoretical framework to generalize the findings from crisis communication case studies. When it comes to the question of how to shape appropriate crisis response strategy for organizations, two dominant theories on crisis communication are Benoit s Image Repair Theory (1997) and Coombs Situational Crisis Communication Theory (Coombs, 1995; Coombs & Holladay, 1996; Coombs & Holladay, 2004; Coombs, 2006). These two theories were designed as roadmaps to understand what strategies are relevant in which situation. However, some argue that a more universal and systemic approach can shape the strategies from an emotion-driven perspective (Jin, Pang & Cameron 2007; Jin, Pang & Cameron 2009). Some researchers argue that, as a key component of crisis perception, emotion involves an interpretation of the stimulus (Carver & Balney, 1977). Lazarus (1991) defined emotion as organized cognitive-motivational-relational configurations whose status changes with changes in the person-environment relationships as this is perceived and evaluated (appraisal) (p. 38). Jin and Cameron (2003) proposed that the importance of emotion in crisis communication was threefold: 18

19 1) [Emotion] as a marker or indicator of the effectiveness of a public relations campaign, with respect to the overall persuasiveness of the press release or issue ad, the appeal of spokesperson of the organization, evaluation of particular organizational claims, and appraisals of other aspects of the public relations practice; 2) As the moderator of impact on a public s attitude toward the organization; 3) As a key factor in organizational decision-making: In dealing with publics in crisis situations, organizational decisions are typically distinguished by several managerial persons being involved in the decision and no one person deciding, which mean to be a shared decision (pp ). How Does Organizational Crisis Affect the Publics Emotion: Understanding the Influencing Process Organizational crisis affects the public s emotion through media messages (Kim & Cameron, 2011). As Coombs (2004) credited, crisis is perceptual, and so the public s perception of crisis was decided by how the media describe it (Kim & Cameron, 2011). The discrete emotions in context, information, and news act as frames in our life, and such frames shape the way people interpret and respond to events in the real world (Nabi, 2003). The Model of Crisis Information Processing (Kim & Cameron, 2011) revealed that people process crisis message in two stages: 1) the initial emotion elicited by crisis news result in people s different depth of thinking, for instance, angry audiences of a corporate crisis would process information heuristically, while sad audiences would develop process news stories systematically, and 2) the initial emotion elicited by crisis news result in people s different attitude towards the corporate; for 19

20 instance, audiences exposed to anger-inducing news tend to evaluate the corporation lower than those exposed to sadness-inducing news. Specifically, the line of scholarship on emotion-based crisis research, which focused on examining how an organizational crisis influence constituencies emotion, found that the publics paid more attention to a human interest frame, for instance, victim relief, than to organization punishment information (Cho & Gower, 2006). Therefore, it was argued that adjusting information that contains a more affective message was helpful in terms of easing the public s negative emotion (Coombs & Holladay, 2004; Kim & Cameron, 2011). Stressful Situations Elicit Negative Emotion Studies have proved that a crisis message would inevitably elicit different negative emotions (Choi & Lin, 2009; Jin, Pang, & Cameron, 2007; Jin, 2009; Jin, 2010; Kim & Cameron, 2011; Jin, Pang, & Cameron, 2007; Jin, Pang, & Cameron, 2009; Jin, Pang, & Cameron, 2012), and scholars have found a range of negative emotions that could be induced from crisis. For instance, Choi & Lin s (2009) research suggested that alertness and confusion were the most frequently expressed attribution independent emotions; psychologist Lazarus (1991) identified six negative emotions in a stressful situation (anger, fright, anxiety, guilt, shame, and sadness); Jin and her colleagues concluded that there are four primary negative public emotions in crisis circumstances: anger, sadness, anxiety, and fright (Jin, Pang, & Cameron, 2007; Jin, Pang, & Cameron, 2009; Jin, Pang, & Cameron, 2012). Crisis experts argued that emotion influences how people think about the organization responsible for a crisis (Kim & Cameron, 2011). One supporting theoretical 20

21 framework to this argument was the appraisal theory, which stemmed from psychology and was applied in strategic communication (Stone & Han, 2007; Kim & Cameron, 2011). Appraisal theory explored emotions that result from people s evaluation, interpretation, and explanation of their surrounding environment (Scherer, Schorr, & Johnstone, 2001). Psychologists viewed emotions as individuals default or automatic response, produced through their subjective evaluation when exposed to events (Scherer, Schorr, & Johnstone, 2001). Among all appraisal theories, the most cited by emotion-focused public relations studies was the Appraisal-Tendency Framework (see Kim & Cameron, 2011; Jin, Pang, & Cameron, 2007). The Appraisal-Tendency Framework, proposed by business scholars, addressed how specific emotions rooted in a prior experience of a similar situation provided reference to make judgments in terms of consumer marketing (Han, Lerner, and Keltner, 2007). Later scholars concluded that the Appraisal-Tendency Framework implies that specific emotion can result in a particular appraisal tendency that influences people s cognitive processing, decision making, and assessment of the situation (Cananaugh, Bettman, & Luce, 2007). For example, fearful people made passive risk assessments while angry people made optimistic risk evaluations (Lerner & Keltner, 2001). Psychologists called people s innate reaction to threatening situation as coping (Stone & Han, 2007). Jin (2010) contributed the psychological concept of coping to establish a crisis communication theoretical framework. Her emotion-based study of people s coping strategy preference in times of crisis evolved from Duhacheck s (2005) psychological study. Duhacheck (2005) proposed that people cope with stressful 21

22 situations with different emotional manifestations, and he developed eight types of psychology processes to categorize people s different coping mechanisms: action, rational thinking, emotional support, instrumental support, emotional venting, avoidance, positive thinking, and denial. Jin (2010) adopted these coping mechanisms. For better understanding, she grouped them into three categories: cognitive (rational thinking and positive thinking), conative (action and instrumental support) and affective (emotional support and emotional venting). Jin suggested that the constituencies innately deal with crisis regarding different levels of predictability and controllability by using three coping strategies accordingly. Emotion and Crisis Responsibility Previous study showed that the public attributed different levels of responsibility when they went through different negative emotions (Lerner & Keltner, 2001; Jin, 2010). In other words, the extent to which the public assigns responsibility to an organization was a significant predictor of their negative emotion (Choi & Lin, 2009; Jin, Pang, & Cameron, 2009). Weiner (1986) argued that when audiences were digesting a piece of information, they were likely to generate general emotion about the crisis. After processing information, audiences would then search for attribution, and such attribution elicited different emotions, for instance anger, fear, surprise, worry, contempt, and relief (Choi & Lin, 2009). Similarly, Jin (2009) held that emotion in crisis was related to crisis predictability and controllability. She examined the variances in the emotional responses of members of the public by testing their appraisal of crisis predictability and controllability. The result suggested that: 1) when stakeholders were exposed to highly predictable crisis 22

23 situations that they had little control over, sadness was reported as the primary emotion; 2) when they were exposed to crisis situations that were both hard to predict and difficult to control, fright was reported as the primary emotion; 3) when constituencies were exposed to highly predictable and highly controllable crisis circumstances, anger was reported as the primary emotion. Coombs and Holladay (2004) also tested the influence of emotions in the SCCT model by integrating crisis responsibility. According to their research, if the public attributed a higher level of crisis responsibility to an organization, they would feel stronger anger and schadenfreude (satisfaction with the pain of other people) (Coombs & Holladay, 2005). The stronger negative emotion members of the public hold, the less likely they would be supportive of an organization (Coombs & Holladay, 2004). Integrated Crisis Mapping (ICM) Model The Integrated Crisis Mapping model was proposed by Jin, Pang, & Cameron (2007; 2009; 2012). It was a systematic model that combined audiences response, corporate action, and crisis type together, which was regarded as a crisis management model mapping out stakeholders hearts. In this model, they categorized crises according to the levels of the publics cognitive coping and levels of the organization s engagement, and predicted the publics negative emotion in a range of crises. The model was initially proposed and named as Integrated Crisis Mapping in the year 2007, and in the year 2012, the authors tested the model with the scientific method and revised it (see Figure 2-1). Four elements were included in the IMC model: primary negative emotions, moderate engagement, conative coping (i.e. publics takes actions to deal with a stressful situation), and types of crises. The main conclusions of this model 23

24 were: 1) crises elicit four primary negative emotions, i.e. anger, sadness, fright, and anxiety; 2) among those four primary negative emotions, anxiety is the default emotion, which means whenever there is a crisis, people s first emotional response is anxiety; 3) the public s innate reaction to crisis (i.e., cognitive processing) lead to external manifestation; and 4) emotions are a reliable indicator in the public s interpretation of crisis situations, and segmented publics are distinguishable by their emotional engagement (Jin, Pang, & Cameron, 2009; Jin, Pang, & Cameron, 2012). The ICM model contributed to mapping out the negative emotions that may occur to stakeholders during crises, and it addressed the significance that an organization strategies should be in accordance with the emotional needs of the public (Jin, Pang, & Cameron, 2007; Jin, Pang, & Cameron, 2009). Emotion-driven crisis communication studies ultimate goal has always being to serve as a roadmap to guide public relations practitioners day-to-day work. Researchers suggested that the attacking-an-accuser, scapegoat, and excuse strategies are the most acceptable to an angry publics, and compensation and apology are preferred when dealing with a sad publics (Jin, 2009); also, people tend to evaluate corporations more positively when the message contains more emotional appeal than information with little emotional appeal (Kim & Cameron, 2011). Attribution of Blame When social psychologists speak of attribution, they often mean responsibility attributions and blame attributions, which are judgments of a moral nature (Malle, 2007). Social psychologists assumed that people develop explanations about what is happening to make sense of the world, and put forward attribution theory to provide an 24

25 insight into understanding how people explain their own and others behavior (McDermott, 2009). Psychologist Fritz Heider first developed attribution theory in his book The Psychology of Interpersonal Relations (Heider, 1958). In the 1970s and 1980s, Bernard Weiner and his colleagues developed Heider s theory; a final stage of development of attribution theory was done by Harold Kelley, who investigated how people validate their perceptions by consistency, distinctiveness, and consensus (Simmering, 2007). There are three dimensions to scrutinize attributions: stability, external control, and personal control/locus (Weiner, 1986). Stability reflects whether the cause of the event is frequent. If the same cause appears repeatedly, it is stable, but the mistake is unstable or it shows infrequently. External control refers to whether the cause of the crisis was controllable or uncontrollable by any other element. For instance, if another element controls what happens to the stakeholder, there is an external control in the situation (Coombs 2007). Locus reflects whether the event s cause is something about the actor or something about the situation (Coombs & Holladay, 2004). Since research on external control and locus demonstrates an overlap, Wilson et al. (1993) suggested that the two causal dimensions be combined as one. Although primarily used to examine interpersonal processing, Attribution Theory has been applied to many social science disciplines (Hart, 2005). For example, Coombs (2006) applied it to crisis management to examine perceptions of crises and their impact on organizational image. He held that people in threatening situations tend to find out who is responsible for leading the crisis; thus, if an organization is blamed by the public, 25

26 its image sustains damage. The next section will focus on Coombs application of Attribution Theory to crisis communication. Situational Crisis Communication Theory Inspired by psychological research on Attribution Theory, Timothy Coombs spent over ten years building his Situational Crisis Communication Theory (SCCT). Because the two key characteristics of crises that they are unexpected and negative (Coombs, 2006) are also the key traits that drive people s need to search for causes of an event in Attribution Theory (Weiner, 1985), it is reasonable to connect crisis with Attribution Theory. The SCCT was primarily built around protecting an organization s image; however, the author took legal and financial cost into consideration (Coombs, 2005). Coombs (2004) held that a crisis is stakeholders perception of the situation; therefore it is essential to identify what factors shape how stakeholders perceive a crisis. Given its origins in Attribution Theory, the central point of the SCCT was the idea of crisis responsibility. Coombs (2007) stated that stakeholders decide whether or not an organization was responsible or not responsible for a crisis, and if stakeholders attribute greater responsibility to the organization for a crisis, the organization s image would be affected negatively. The SCCT was a two-stage system: first, identify crisis type; second, identify crisis intensifiers. There were three types of crisis in accordance with three levels of attributions of organizational crisis responsibility. The first crisis cluster was victim, which means the organization has very low responsibility, or sometimes even the organization itself is a victim of the crisis. This cluster included natural disaster, rumor, workplace violence, and product tampering crises (Coombs & Holladay, 2007). The 26

27 second one was the accidental cluster. The defining characteristic of an accidental crisis was that it was unintentional. This cluster included challenges, megadamage, technical breakdown accidents, and technical breakdown product recalls (Coombs, 2006). This cluster produced moderate attribution to the responsible organization. The third cluster, and the most difficult one to manage, was called preventable crisis. In a preventable crisis, stakeholders often strongly believe that the organization is responsible for a crisis. Some examples for this type of crisis included three variations of organizational misdeeds, human breakdown product recall, and human breakdown accident (Coombs, 2006). In addition to crisis type, Coombs (2004) and his colleague added intensifiers to predict stakeholders perceptions of a crisis. Intensifiers are variables that can alter attributions generated by crisis type (Coombs, 2006, p. 182). There were three intensifiers in SCCT: 1) crisis history, or whether a similar crisis happened in the past; 2) crisis severity, or the amount of damage from the accident; 3) and relationship history, or whether the organization had a good record of behavior toward stakeholders (Coombs, 2004). So far, studies have proved that similar crises reduce the publics perception of organizational responsibility, while a negative performance history intensifies the attribution of crisis responsibility (Coombs, 2004; Coombs, 2006). When applying crisis situation theory to practical work, Coombs (2004) suggested that practitioners should better match crisis response strategies with a combination of assessments of crisis type and modifiers. The basic idea behind this guidance was that as stakeholders attribution of blame intensifies, the organization should accept a higher level of responsibility (Coombs, 2006), and that the objective of crisis response strategy 27

28 was to shape attributions of the crisis, change perception for the organization in crisis and reduce the negative affect generated by the crisis (Coombs, 2007). SCCT developed four postures of crisis response strategies, based upon perceptions of accepting responsibility for a crisis. Four primary types of crisis response strategies were proposed: 1) denial, 2) diminish, 3) rebuild, and 4) bolster. Table 2-1 explains the SCCT crisis response strategies (Coombs, 2007). Coombs (2006) proposed that different crisis response strategies imply different degrees of acceptance of crisis responsibility. For example, victim organizations should use denial. In an accident crisis, organizations were encouraged to use the diminish strategy, while in a preventable type of crisis an organization should use the rebuild strategy. The bolster strategy was found to be helpful in every situation; therefore, it was argued, bolstering should be used together with other strategies to optimize effectiveness (Coombs, 2006). Despite SCCT s status as a leading theory in the crisis communication field, critics refuted that certain public relations crisis situations were better settled if organizations adopted specific crisis response strategies. For example, in the real world, the Red Cross only chose crisis response strategies suggested by SCCT one-third of the time (Sisco, Collins, & Zoch, 2010). Claeys and his colleagues (2010) also learned that the interaction between the SCCT crisis type and the crisis response strategy to deal with organizational reputation was not that significant. In addition to that, evidence has shown that crisis communication outside of America was practiced not exactly the same with American public relations professionals (Harro-Loit, Vihalemm, & Ugur, 2012). 28

29 Cultural Issues in Crisis Communication While the U.S. is the birthplace of public relations theory and has dominated public relations for several decades, intercultural/international public relations is one of the most quickly evolving areas in public relations, and U.S. dominance in this area is fading (Botan & Hazleton, 2006). The rapid growth of public relations practice worldwide brings us many new cultural insights (Botan & Hazleton, 2006). With the growing number of multinational corporations and the development of modern media, crises can travel across geographical boundaries easily (Lindholm & Olsson, 2011). Despite the fact that intercultural public relations is still a underdeveloped area (Boton & Hazelton, 2006), scholars argued that there has been an increasing need for public relations practitioners to develop an understanding of not only different cultures, but also how cultural variances influence modern public relations work (Kanso, Sinno, & Adams, 2001). Intercultural crisis experts often look at differences of practice through the lens of Hofstede s cultural dimensions to scrutinize and analyze cultural differences (An, Park, Cho, & Berger, 2010; Haruta & Kirk, 2003; Wertz & Kim, 2010; Taylor, 2000). Cultural dimensions established criteria for describing different cultural features and provided an intercultural perspective to crisis management. Social psychologist Geert Hofstede (1980, p. 25) defined culture as the collective programming of the mind which distinguishes the members of one human group from another Culture, in this sense, includes systems of values; and values are among the building blocks of culture. Hofstede (1984) and his colleagues examined value differences among national societies and identified five areas of common differences. 29

30 The differences are identified as power distance, femininity versus masculinity, uncertainty avoidance, long-term versus short-term orientation, and individualism and collectivism (Hofstede, 1984). Hofstede then investigated how these cultural variances influenced corporate behavior in many countries. Among all five cultural dimensions, collectivism and individualism is the one that has been studied most thoroughly, especially in cross-cultural psychology, to explain differences between Asian culture and Western culture (Triandis, 1995). This dimension describes the relationship between the individual and the collectivity which prevails in a given society (Hofstede, 1984, p. 213). Individualism emphasized the self, emotional independence, self-initiative (Hofstede, 1984), self-dependency, freedom of choice, and individual rights over duties (Bellah, Madsen, Sullivan, & Tipton, 1985), whereas collectivism addressed the group s interest over that of the individual, dependency on group, conformity, harmony, sharing duties, and obligations (Hofstede, 1984). Power distance is a cultural variability dimension that concerns the extent to which people accept an unequal distribution of power (Martin & Nakayama, 2008, p. 109). For example, North European countries value small power distance (Taylor, 2000), and citizens from those countries believe that less hierarchy is better and that power should be used only for legitimate purposes (Martin & Nakayama, 2008). Taylor (2000) held that power distance can help public relations practitioners better understand the dynamics of organization-public relationships in international context (p. 279), because it reflects how people deal with inequality and conflict. Also, it was argued that people from regions with a higher power distance were less likely to trust authority and power. 30

31 Masculinity-femininity refers to the concerns [that] the degree of being feminine valuing fluid gender roles, quality of life, service, relationships, and interdependence and the degree of being masculine emphasizing distinctive gender roles, ambition, materialism, and independence (Hofstede 2012). He further explained that the masculine side of this dimension stands for a society that prefers achievement, heroism, assertiveness, and material reward for success. A society with higher level of masculinity is more competitive (Hofstede, 2012). Uncertainty avoidance concerns the extent to which uncertainty, ambiguity, and deviant ideas and behaviors are avoided (Martin & Nakayama, 2008, p. 104), which means that people feel threatened by ambiguous situations and try to establish more structure to compensate for the ambiguity. Great Britain, Sweden, Hong Kong, and the United States are all weak uncertainty avoidance societies, and people from those regions prefer to limit rules, accept dissent, and take risks (Martin & Nakayama, 2008). Long-term versus short-term orientation, also known as Confucian dynamism, was developed later than the previous four dimensions in order to compensate for Western bias (Martin & Nakayama, 2008). People with this long-term trait are concerned with the demands of virtue and are more willing to accept slow results (Haruta & Kirk, 2003). Hofstede s cultural dimensions, according to some scholars, have been proved by decades of research and helped reduce uncertainty in intercultural communication encounters (Olaniran, 1996; Martin & Nakayama, 2008). Despite being cited in many fields, the cultural dimensions theory was criticized for being static and not taking social changes, power, diversity, and other activism into scope (McSweeney, 2002). Those 31

32 deficiencies might not have enforced comprehensive understanding of each specific culture, but reinforced national stereotypes (McSweeney, 2002). Cultural Issues, Blame of Attribution, and Negative Emotions Drawing upon Hofstede s (1980; 1984; 2012) five cultural dimensions, crosscultural crisis communication scholars studied several intercultural cases by contrasting organizational responses of different cultures, for instance, a Korean rotten dumpling crisis and an American E. coli spinach crisis (Wertz & Kim, 2010), a Japan Airlines crisis and an American Delta Air crisis (Haruta & Kirk, 2003), and Coca-Cola crises in six European countries (Taylor, 2000), etc. All of these studies revealed significant differences in the use of apology, media strategies, and litigation concerns. For instance, Taylor (2000) concluded from the Coca-Cola crisis in Europe that publics who live in nations that are high in uncertainty avoidance and power distance tend to react more strongly, and more quickly, to perceived threats (p. 277). Cultural variances were found related to stakeholders acceptance to responsible organizations crisis response strategies. Lee (2010) studied organizational crisis in Hong Kong and concluded that apology, compensation, and corrective action demonstrated a hierarchy of acceptance of crisis responsibility, and Hong Kong respondents did not show much sympathy towards apology strategy. This result was in synch with Quattrone s (1982) argument. Quattrone was an American psychological scholar who argued that publics usually did not have a personal understanding of crisis background information and were not able to engage backward chaining. Therefore, acceptance of responsibility, for instance through apology, could be effective in American culture (Weiner, Graham, Peter, & Zmuidinas, 1991). The difference of 32

33 people s acceptance of organizational responsibility indicates publics different level of blame attribution (Coombs, 2010). Thus stakeholders cultural backgrounds affect their attribution of blame. Cultural issues also have impact on people s emotions during crisis. An et al. (2010) compared Koreans and Americans post-crisis reaction, and found that Koreans revealed a higher level of anger than Americans towards the responsible organization. Because Koreans strong collective value, which is symbolized as Cheong, Korean respondents consider the employee as a member of the family, and the organization as an extended family (An et al., 2010, p. 230) so they might regard blaming an individual in an organizational crisis as inappropriate. Hypotheses and Research Questions Based upon the literature review and exxisting theory, the Chinese and American cultures were selected as the two comparative cultures in this study, because the U.S. was the birthplace of public relations as well as crisis communication, and represents Western culture. On the other hand, China, where crisis communication evolved rapidly in recently years, represents Eastern culture, and provides a contrast with Western culture. The following Figure 2-2 demonstrates Hofstede s cultural dimension variances between China and America. According to An et al. (2010), people with different cultural backgrounds would perceive the same organizational crisis differently. They tested the relationship between individualism-collectivism dimension and people s perception of an organizational crisis, and reported that the respondents form a high collectivistic culture had more negative attitudes, emotions, and impressions toward the organization than those respondents 33

34 from a highly individualistic culture (p. 235). As a collectivistic country, China s individualism index scored 20 in Hofstede s (2012) cultural dimensions index, and was close to South Korea s score, 18. Therefore, the author of this paper predicted that Chinese people would response similarly to South Korean in an organizational crisis when compared with American respondents assign more blame attribution and express more negative emotion. Thus from an intercultural and publicbased crisis communication perspective, the author predicted that it would be more difficult to mitigate both Chinese respondents negative emotion and attribution of blame in an organizational crisis. Two hypotheses were proposed: H1: Organizational crisis response strategies would be less effective in alleviating publics attribution of blame in China than in America. H2: Organizational crisis response strategies would be less effective in mitigating negative emotion in China than in America. Previous intercultural crisis communication studies have also suggested other cultural dimensions might be valuable as a tool to analyze cross-cultural differences, and were underscored in previous studies (Haruta & Hallahan, 2003). Confucian dynamism, also known as long-term versus short-term orientation, was found highly related to organizations response in a crisis (Taylor, 2000). Wertz and Kim (2012) also held that those nations with higher levels of Confucian dynamism placed an emphasis on virtue over truth. As such, individuals from countries with a higher score in Confucian dynamism held less tolerance for ambiguity. This mindset makes it more difficult for companies to satisfy stakeholders through avoidance of responsibility. Based on this, the following hypothesis was drawn: 34

35 H3: A higher level of power distance is positively related to respondents higher level of blame attribution and lower level of negative emotion. In regards to crisis communication strategies, Coombs (2006) found that the most frequently used crisis communication strategy types were: denial (taking responsibility for one s action), deal (take actions such as bolstering, praising others, compensation, apology and corrective action), diminish (the attempt of minimizing the responsibility), and silence (not responding to crisis). An, Gower and Cho s (2011) study analyzed organization s frequency of usage of those crisis response strategy clusters, and the breakdown percentage was: denial (47.4%), diminish (9.2%), deal (41.3%), and silence (1.8%). This study will be only focusing on testing the effects of denial, diminish and rebuild clusters. Previous research conducted has shown that silence was more effective in Eastern countries than in Western countries (Lee, 2004; Davis & Holtgraves, 1984). Easterners demonstrated a more tolerant attitude towards non-comments response due to the fact that, in their culture, a silent, reserve gesture is often regarded as an act of wisdom. This can be attributed to the Confucius mindset of think three times before an action (Bond, 1991). Secondly, some crisis response strategies under deal cluster have also been tested in cross-cultural studies, while some were not. For example, An, Park, Cho, and Berger (2010) suggested that organizations from highly collectivistic culture should not punish individuals in an internal crisis. Corrective actions such as punishing staff could elicit negative public sentiment toward the company. However, the cross-cultural effectiveness of rebuild strategies, such as compensate were seldom studied. 35

36 Other strategies, such as bolstering and praising others, were considered as secondary strategies (Coombs, 2007). These strategies attempt to show goodwill, caring for victims, and sympathy towards stakeholders (Coombs, 2007). Coombs (2007) believed that this type of crisis response strategy were used to supplement the primary responses (e.g. diminish, denial and rebuild), not as a replacement, and were considered effective in any type of crisis communication. Since research has been conducted in the aforementioned areas, this study is designed to test the effectiveness of these three clusters in cross-cultural crisis communication. 36

37 Figure 2-1. Revised Integrated Crisis Mapping Model Jin, Pang, & Cameron, Toward a public-driven, emotion-based conceptualization in crisis communication: Unsearthing dominant emotions in multi-staged testing of the integrated crisis mapping (ICM) model,

38 Table 2-1. Crisis Response strategies, by Postures Strategy Definition Denial Posture Attacking the Accuser Denial Scapegoating The crisis manager confronts the person or group that claims that a crisis exists. The response may include a threat to use force (e.g., a law suit) against the accuser. The crisis manager states that no crisis exists. The response may include explaining why there is no crisis. Some other person or group outside of the organization is blamed for the crisis Diminishment Posture Excusing Justification The crisis manager tries to minimize the organization s responsibility for the crisis. The response can include denying any intention to do harm or claiming that the organization had no control of the events that led to the crisis. The crisis manager tries to minimize the perceived damage associated with the crisis. The response can include stating that there were no serious damages or injuries or claiming that the victims deserved what they received. Rebuilding Posture Compensation Apology Bolstering Posture Reminding Ingratiation The organization provides money or other gifts to the victims The crisis manager publicly states that the organizations takes full responsibility for the crisis and asks for forgiveness. The organization tells stakeholders about its past good works. The organization praises stakeholders. Victimage The organization explains how it too is a victim of the crisis. 38

39 - Figure 2-2. Contrast China s and America s Five Cultural Dimensions. 39

40 CHAPTER 3 METHODOLOGY An experiment is an appropriate method to test the hypotheses for this study. Previous studies showed that an experiment was the proper way to test, compare, and contrast differences between cultures (An, Park, Cho, & Berger, 2010; Diener, Osshi & Siswas-Diener, 2012). Therefore, this study adopted a 2 (nationality: China vs. the U.S.) 4 (crisis response strategies: denial, justification, apology, and control group) between-subjects experiment design to test the hypotheses. Participants This study collected a total of 330 responses, but after excluding incomplete responses 16.97% (n =56), a total of 274 were participated in this study. Among the participants, 48.5% (n=133) were Chinese while 51.5% (n=141) were American respondents. The Chinese respondents were recruited at Remin University of China, whereas the American participants were recruited at the University of Florida. Since it is difficult to reach Chinese respondents via traditional paper-pencil method, this study conducted this experiment online via Qualtrics. Gender (male n = 99, 36.1%; female n = 175, 63.9%) and age (M = 21.5, SD = 2.53) were measured as possible variables. Procedure All participants were asked to report nationality and power distance level before exposing to stimuli. Then, they were randomly assigned to only one of the experimental conditions. American treatment groups were exposed to an English crisis news article in addition to three organizational crisis response strategies, while Chinese treatment groups were exposed to the same translated stimuli. After reading the crisis news article 40

41 participants were asked to report their attributions of blame and negative emotions toward the company. Stimulus Material Crisis type This study used a preventable crisis type a technical error accident leading to cell phone A preventable crisis type was chosen due to its frequency, since 85% of organizational crises were of the preventable type (An, Gower, & Cho, 2011). Cell phone explosion was chosen, because: 1) a cell phone is a product in which students would be highly invested, and participants will perceive a daily used product more seriously (Coombs, 2007); and 2) a cell phone problem is a scenario that respondents are familiar with, because similar crises have happened before. An example of this was the Kyocera Wireless case in 2004 (Charny, 2004). A fictitious company was used to prevent any previous company judgments that might be attributed to actual organizations. The constructed crisis news article was about a cell phone explosion at (i.e., fictitious company), caused by human error. The news article includes a brief description about the explosion and an organizational crisis response (i.e., denial, justification, or apology). In order to rule out possible confounding effects, the location of the crisis was kept domestic: the explosion crisis happened in Hebei Province for Chinese participants, and this province is adjacent to Beijing, where the majority of Chinese respondents live; and Florida for American participants, where American respondents were recruited. For the same concern, the choice of news agency was also kept domestic: Associate Press for American participants, and Xinhua Press for Chinese respondents. 41

42 The company s name remained the same, but was translated into Chinese for Chinese participants. In order to help participants believe the news stories were real, the experiment was conducted online with news stories. In addition, all stimuli were translated into Chinese for Chinese participants by the author, and were verified by a Chinese doctoral student. A pretest was conducted to ensure that participants response would not be affected by readability of translation. Participants were asked to answer I think this news article reads well in a seven-point Likert scale (1 = dis, 7 = strongly ). Respondents reported similar evaluations in terms of the stimuli s readability (F (1, 69) =.326, p =.570). Crisis response strategy Although An, Gower, and Cho (2011) categorized four crisis response postures (denial, diminish, rebuild and bolstering), some other scholars suggested silence to be a fifth posture (see Lyu, 2011). This study, however, did not examine the silence posture, because previous studies have suggested that the silence posture is more effective in Eastern countries. Also, this study did not examine secondary strategies such as bolstering, because Coombs (2007) stated that this type of crisis response strategy would be more effective as a supplemented strategy, while the other three types of strategies are not replaceable. Since denial, diminish and rebuild strategy clusters include more than one specific crisis response strategy (see Table 2-1), this study selected only one strategy from each posture: a denial strategy from the denial posture, a justification strategy from the diminish posture, and an apology strategy from the rebuilding posture. A denial strategy 42

43 was selected because it has been most often used in preventable crisis, and 47.7% of organizational responses to a preventable crisis were denial (An, Gower, and Cho, 2011). A justification strategy was selected because the strategy was frequently used in the Eastern culture, but seldom seen in the Western culture (e.g., Wertz & Kim, 2010). The salient difference in the adoption of justification would provide a better change to explore cultural differences in stakeholders attribution of blame and negative emotions inflicted by the crisis. Finally, previous studies have shown that in some Eastern cultures, such as Korea and Japan, apologizing does not necessarily mean accepting responsibility (Wertz & Kim, 2010; Haruta & Kirk, 2003). Thus, to explore further cultural differences in emotions and blame attributions, this study selected an apology strategy from the rebuilding posture. For a denial strategy, the news article included crisis responses such as spokesperson says this incident was unrelated to the. The justification strategy condition included the company responses of considering that a high explosion rate in other advanced new technology products, s products are still safe, especially compared to contemporary products. For an apology strategy, the company issued an apology offering deep and sincere condolence for the victims of the crisis. In order to insure the appropriateness of those strategies, all three response messages were reviewed and verified by a crisis communication expert. 43

44 Measures Power distance scale Power distance levels of respondents were measured because this study is interested in the relationship between respondents power distance level and their attributions of blame to the company in a crisis. Power distance is a measure of interpersonal power or influence between [boss] and [subordinate] as perceived by the least powerful of the two (Hofstede, 1984, p ). Hofstede (1984) developed several questions to test the power distance dimension. To measure participants individualism, this study adopted and revised Klein & Dawar s (2004) power distance scale, such the following three questions measured through Likert scales ranging from 1 (strongly dis) to 7 (strongly ). 1. Students should not dis with decisions made by professors. 2. I believe that students should not treat teachers as equals. 3. Teachers are expected to take all initiatives in class. Negative emotions To measure respondents level of negative emotions generated by the crisis, this study adopted emotion measures from Jin, Pang, and Cameron (2007; 2009; 2012). Participants will be asked to respond to the organization story made : 1) angry, irritated, annoyed; 2) sad, downhearted, unhappy; 3) scared, fearful, afraid; and 4) nervous, anxious, worried. Responses were recorded with a 7- point Likert-type scale, with 1 being strongly dis and 7 being strongly. Blame attribution 44

45 This study adopted and modified Klein and Dawar s (2007) scale to measure consumers blame attributions. It is a 7-point Likert-type scale that asks respondents to address the following: 1) is responsible for the 2) should be held accountable for the 3) The explosion incident is the fault of. Measures reliability The reliability test was performed to examine the negative emotions scale, attribution of blame scale and power distance scale. The data showed that for negative emotion, Cronbach s alphas for the four negative emotion items was.83, for the three negative attribution items was.86, and the three power distance items score were.995. Therefore the measures in this study are reliable. 45

46 CHAPTER 4 RESULTS Manipulation check To ascertain whether the experimental manipulations were effective, a one-way ANOVA were performed. A significant difference among the three crisis response strategies was found. For denial response strategy, those who exposed to the denial response strategy considered the company denied their responsibility for the crisis (M = 5.14, SD =1.700) more than those in the apology condition (M = 3.101, SD =1.619, Tukey HSD p <. 01), but not significantly different from those in the justification condition (M = 4.75, SD =1.631, Tukey HSD p =.99). Respondents exposed to justification response strategy considered the company justified the responsibility for the crisis (M = 4.17, SD = 1.840) more than those in apology strategy condition (M = 2.61, SD = 1.583, Tukey HSD p <.01), but considered it not much different from the denial response strategy (M = 3.30, SD = 1.727, Tukey HSD p =.317). Participants exposed to apology strategy considered the company apologized for the crisis (M = 5.22, SD = 1.57) more than those in justification condition (M = 2.55, SD = 1.481, Tukey p <.01) or denial condition (M = 2.53, SD = 1.657, Tukey p =.039). Considering in a preventable type crisis, people may regard justification as a kind of denial, because both of the strategies were aiming at evading responsibility, therefore it is acceptable that people consider these two strategies are similar in this case. Therefore, the stimuli manipulation was successful. Test of Hypotheses H1: Organizational crisis response strategies would be less effective in alleviating publics attribution of blame in China than in America. 46

47 H1 examined the effectiveness of different crisis response strategies in terms of mitigating Chinese and American audiences attribution of blame. Because all three items in the attribution of blame measurement were testing the same dimension, the author of this paper collapsed the value of those three items. An ANOVA test was conducted to test this hypothesis. Results showed that the difference between Chinese and the US participants attribution of blame was significant, F (1, 266) = 4.15, p =.04. As predicted in H1, organizational response strategies were more effective among the US respondents (M = 5.51, SD = 1.06) than Chinese respondents (M = 5.78, SD = 1.03) in terms of alleviating attribution of blame. Thus, H1 was supported. Results also indicated that there was a significant main effect of organizational strategies in terms of mitigating respondents attribution of blame: F (3, 266) = 3.72, p =.01. When looking at each individual corporate response s effectiveness, denial (M = 5.32, SD =.216) turned out to be the most effective corporate response strategy among 3 treatment groups and 1 control group, and apology (M = 5.82, SD =.122) has the weakest effect in terms of mitigating respondents attribution of blame. Significant difference existed between denial (M = 5.32, SD =.216) and the control group (M = 5.83, SD =.122, Tukey HSD p =.024). A considerable difference also exists between apology (M = 5.82, SD =.123) and denial (M = 5.32, SD =.126, Tukey HSD p =.045). Data also revealed that there is a significant two-way interaction effect between nationality and crisis response strategies in terms of blame attribution: F (3, 266) = 3.09, p =.028. When examining the data for each treatment group, Chinese participants attributed significant higher blame levels to the company than Americans for justification (p =.05; 2 p =.014) and apology (p =.04, 2 p =.016). However, the participants from 47

48 both countries who exposed to control stimuli (p =.087, 2 p =.011) and denial stimuli (p =.108, 2 p =.010) did not revealed much difference in terms of blame attribution. Please see Table 3-1 and Figure 3-1. Among all 8 experiment groups, Chinese apology treatment group revealed the highest level of blame, while the denial strategy in China created the lowest level of blame attribution. H2: Organizational crisis response strategies would be less effective in mitigating negative emotion in China than in America. A MANOVA test was conducted to explore H2, which predicted a significant main effect of nationality in terms of crisis response strategy s impacts on negative emotions. All four items of negative emotions were explored using a MANOVA because this study interested in crisis strategies impact on different emotions, such as anger, sadness, fear and anxiety. The effect of nationality on negative emotion was significant (Wilks Lambda =.947, F (1, 266) = 3.704, p =.006), indicating Chinese participants tended to reveal more negative emotions than American participants. When examining nationality s effect on each individual negative emotion, results showed that anger (F (1, 266) = 5.12, p =.022) and anxiety (F (1, 226) = 12.40, p =.001) was more significantly affected by nationality than sadness (F (1, 266) = 1.85, p =.175) and fear (F (1, 266) = 3.29, p =.071). When comparing people from which country has more intensive negative emotion, Chinese respondents experienced significant higher anger (Mean Difference =.371, SD =.161, p =.022, 2 p =.022) and anxiety (Mean Difference =.632, SD =.179, p =.001, 2 p =.001) than American respondents (see Table 3-2). However Chinese 2 participants sadness (Mean Difference =.242, SD =.178, p =.175, p =.007) and fear 48

49 (Mean Difference =.330, SD =.182, p =.071, 2 p =.012) was not significantly different from American respondents. Therefore, for H2, Organizational crisis response strategies are less effective in mitigating anger and anxiety in China than in America, but have similar effect in mitigating sadness and fear in China and America. Corporate response strategy s impact on participants negative emotion was also significant (Wilks Lambda =.796, F (3, 266) = 5.23, p <.0001). There were significant differences among different crisis strategies in terms of all negative emotions: anger: F (3, 266) = 10.28, p <.0001; sadness: F (3, 266) = 4.90, p =.002; fear: F (3, 266) = 12.95, p <.0001; anxiety: F (3, 266) = 11.31, p < When the company used an apology strategy, respondents regardless of nationality tend to show less anger than control group (Tukey HSD, p <.0001), justification group (Tukey HSD, p <.0001), and denial group (Tukey HSD, p <.0001). Also, the experimental groups exposed to justification strategy (Tukey HSD, p <.0001), apology strategy (Tukey HSD, p <.0001), and denial strategy (Tukey HSD, p =.003) displayed less fear than the control group. Respondents from justification group (Tukey HSD, p =.001), apology group (Tukey HSD, p <.0001), and denial (Tukey HSD, p =.004) revealed less anxiety than the control group. However, no significant interaction effect was found between nationality and crisis response strategies in terms of negative emotions: Wilks Lambda =.954, F (3, 266) = 1.03, p =.415. H3: A higher level of power distance is positively related to respondents higher level of blame attribution and lower level of negative emotion. 49

50 H3 predicted power distance s correlation with two variables: level of blame attribution and level of negative emotion. Mean scores of power distance measure were calculated for an ANOVA test. Surprisingly, the ANOVA test revealed that there was no significant difference in the level of power distance between Chinese and Americans, F (1, 266) =.107, p =.744. A series of correlation tests were performed to identify the relation between power distance and blame attribution and between power distance and negative emotion. The data showed that there is no correlation between power distance and attribution of blame (Pearson Correlation α =.073, p =.114). The analysis run between power distance and negative emotion yielded the same result: anger: Pearson Correlation α =.117, p =.027; sadness: Pearson Correlation α = -.002, p =.484; fear: Pearson Correlation α =.110, p =.034; Anxiety Pearson Correlation α =.110, p =.035. Therefore, H3 was not supported. 50

51 Table 4-1. Mean and Standard Deviation of Blame attribution in China and the US Nationality Corporate Response M SD China Control Justification Apology Denial The US Control Justification Apology Denial Figure 4-1. Interaction between Nationality and Corporate Response in Reducing Publics Attribution of Blame 51