DEFENDING YOUR COLLECTIVE FACE IN FRONT OF CO-NATIONALS AND THE PATRIOT AND THE TRAITOR: FOREIGNERS By Yi Zhu A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Communication—Doctor of Philosophy 2019 ABSTRACT DEFENDING YOUR COLLECTIVE FACE IN FRONT OF CO-NATIONALS AND THE PATRIOT AND THE TRAITOR: FOREIGNERS By Yi Zhu The current between-subjects study investigated how participants from two cultural groups (100 American domestic students and 115 Chinese international students) deal with threats to their collective face elicited from a critic who is either an ingroup member, an outgroup member, or an identity-unspecified member in an intercultural-communication context with a laboratory experimental design. Chinese students reported higher collective face concerns and lower liking towards a person who criticized their collective face compared with Americans. While encountering criticism targeting their countries, Chinese felt higher discomfort feelings compared with Americans. Chinese participants’ discomfort feelings in the ingroup-critic condition were more influenced by their collective face concerns compared with Americans in the same condition. The practical and methodological implications of this study were also discussed. This dissertation is dedicated to my parents, Che Zhu, and Yuzhu Suo. Thank you for supporting me. iv ACKNOWLEDGEMENTS I want to express my thanks to everyone who helped with this project. I started to collect the data since Spring semester in 2017 and just defended this dissertation in June 2019. It took a little bit longer than I expected before while the results are interesting enough for further investigations. I would like to thank my advisor, Professor Mary J. Bresnahan, who offered me so much advice and encouragement during my entire studying experience in Michigan State University since 2011. I cannot complete this project without you! I am so glad to be your advisee working with you and the rest of our research team members! As noted in an old Chinese saying, the best instructor prefers being a role model for students rather than teaching students through verbal instructions. I learned so much from you to be a committed intercultural communication scholar. I want to thank Professor Franklin J. Boster for your suggestions to foster this project. You helped me so much with your research-method expertise! I also want to thank Professor James W. Dearing for your ideas to inspire this project and support with my data collection. I would like to thank Professor Adan Quan for all of your feedbacks to improve this project. I want to acknowledge Office for International Students and Scholars, Office of the Registrar, and English Language Center at Michigan State University for their collaboration and help with the promotion of my project. I want to express my thanks for Chinese Student & Scholar Association (CSSA), Chinese Undergraduate Student Association (CUSA), and all other Chinese international students' communities which helped me to promote my study at Michigan State University. I also want to express my special thanks for all of my participants in this study, my research assistants, Ian P. Anastasia, Benjamin L Wilson, Rongjia Yan, Ben Prchal, Brandon v M. Walling, Nicholas D. Bahoric, Guiya Lyu, Evan J. Leslie, Alex Dickman, Zhiqing Zhou, Siyuan Ma, and Henry A. Goble, my students and friends who helped me to promote this study among their friends, and those Comm. Dept. faculty members who allowed me to promote my study at their classes. I also want to express my thanks for my parents for your love and emotional and economic support with me. My dissertation was supported with Michigan State University's Graduate School Dissertation Completion Fellowship and the funding from the Department of Communication at Michigan State University. vi TABLE OF CONTENTS LIST OF TABLES…………………………………………………………………………… viii INTRODUCTION……………………………………………………………………………… 1 LITERATURE REVIEW……………………………………………………………………….. 2 METHOD…………………………………………………………………………………… 12 Participants and Procedure……………………………………………………………….12 Measures………………………………………………………………………………... 16 Collective face concerns……………………………………………………… 16 Liking towards the critic……………………………………………………… 16 Liking towards the noncritic witness…………………………………………… 17 Positive and negative affect towards the study…………………………………. 17 RESULTS………………………………………………………………………………………. 17 Other Results……………………………………………………………………………. 23 DISCUSSION………………………………………………………………………………… 25 Cultural Differences and Traitor Effect………………………………………………… 25 Implications…………………………………………………………………………… 27 Practical implications…………………………………………………………… 29 Methodological implications…………………………………………………… 30 Limitations……………………………………………………………………………… 32 CONCLUSION……………………………………………………………………………… 36 APPENDICES………………………………………………………………………………….. 38 APPENDIX A: Sample Scripts………………………………………………………… 39 APPENDIX B: Measures………………………………………………………………. 45 REFERENCES…………………………………………………………………………………. 49 vii LIST OF TABLES Table 1: Measure Indices……………………………………………………………………… 46 Table 2: Descriptive Statistics for Conditions………………………………………………….. 47 Table 3: Ordinary Least Squares Regression Results to Predict Feelings of Discomfort……… 48 viii INTRODUCTION This study is the next step in investigating a newly proposed concept --- collective face (Zhu, 2014; Zhu & Bresnahan, 2018) --- in an intercultural-communication context. Collective face is defined as an individual’s concern with presenting a collective positive image of his or her salient community in the eyes of outgroups. A person with high concern for collective face will try to maintain, protect, and defend collective face so members of outgroups will not disparage his or her community. This study investigates four questions. The first question is whether and how people in different cultures respond to threats directly targeting their collective face in an intercultural communication context. The second question considers whether people in different cultures respond differently to collective-face threats elicited from members of the ingroup versus outgroups. The third question asks how people in different cultures respond to noncritical outgroup members who merely witness the context where people’s collective face is threatened. The final question asks whether there is any cultural difference regarding people’s positive and negative affect due to threats to collective face. Two causal factors of collective face investigated in this study include: 1) the cultural identities of participants, and 2) whether or not the membership of the critic eliciting collective face threat makes any difference. Cultural identity aligns with the country of origin whereas ingroup-outgroup membership pairs participants with either a critic from the same cultural background (the ingroup-critic condition), a critic from a different culture (the outgroup-critic condition), and a critic with an unspecified identity (the control condition) to see whether this influences participants’ attitudes and affect. To this point, a few studies have tested the viability of the collective face construct in an intercultural communication context. For example, Zhu and Bresnahan (2018) found a 1 significant cultural difference between American domestic students and Chinese international students regarding their collective face concerns showing that Chinese international students were more sensitive to their group image in front of the outgroup compared with American domestic students. Collective face is conceptually related to cultural values (Zhu, 2014) since cultural values such as Chinese Confucianism endorsing extended selfhood (see Ho, Peng, Lai, & Chan, 2001 for details) and group cohesiveness are likely to promote collective face. Collective face is also related to the ingroup versus outgroup status (see Hogg, Terry, & White, 1995 for a detailed discussion about ingroup and outgroup) since collective face represents an individual’s desire for the ingroup to have a positive evaluation from members of outgroups. In an earlier study conducted by the author (Zhu & Bresnahan, 2018), the ingroup versus outgroup status was manipulated through asking participants to imagine scenarios about interacting with a close friend who either shares the same national identity with the participants or not; however, little evidence about the association between collective face and ingroup- outgroup membership was found. In this earlier study, participants did not report strong collective-face in response to imagined scenarios. This result suggests that simply asking participants to imagine witnessing a co-national experiencing a face threat may not be a sufficiently strong research manipulation to elicit a collective face response. The current study conducted an experiment intended to induce a direct threat to the participants’ own collective face to test whether it matters if someone from the ingroup or the outgroup delivers a collective face threat. LITERATURE REVIEW Goffman (1967) brought the concept of face from Chinese culture to the West focusing on the positive social value of face and its self-presentation functions. Brown and Levinson 2 (1987) extended the discussion of face to the subtle nature of face-threatening acts (FTAs), including positive face which is the need for being approved and appreciated, and negative face which is the need for freedom from any imposition. However, most previous face-related studies investigated either an individual speaker or addressee’s positive or negative face (e.g., Cai & Wilson, 2000; Park & Guan, 2006) while a macro level analysis of face as it extends to a whole group of people has been only minimally examined. While Goffman (1967) described that face can be shared with others, he did not describe whether face could be collectivized or shared within a larger community. In contrast to conceptualization and study of face in the West, Chinese scholars have conceptualized face as a macro concept (Hwang, 2012; Zhai, 2011). Ho et al. (2001) discussed underlying Asian identity values of “self-in-relations” in contrast to the Western self-other demarcation. This suggests that the concept of self in Asian societies is based on the collectivized self in terms of its relation with others. Hwang (2012) similarly described that Chinese are motivated for the face of the greater self and that Confucian relationalism makes an individual’s relational others share this individual’s feelings too. Hu (1944) discussed that Chinese are afraid of losing their country’s face in front of foreigners. Zhai (2011) discussed several examples about Chinese athletes who worried about losing face for their country and people when they failed to win medals. Hail (2015) similarly observed Chinese international students’ national identity become more salient due to being abroad and they have a tendency to identify with the Chinese nation-state and show their loyalty to China when they have conversations about China with Americans. These examples imply that, for Chinese thinkers, face is conceived as positive social values that are shared within a large population. People within this population are concerned with their own behaviors so they will not harm their 3 collective image in the outgroup members’ eyes. In this view, collective face is clearly important to Chinese scholars and appears to operate apart from Brown and Levinson’s (1987) face framework in which positive face is defined as one’s desirable social identity and negative face defined as concern about intrusions on personal autonomy. Collective face is different from both positive face and negative face insofar as collective face is beyond a dyadic level of analysis. Collective face is not concerned with the desire for social approval for a single person or personal autonomy. It is also clear that Chinese also have micro positive and negative personal face examined for Chinese in previous studies (e.g., Oetzel et al., 2001; Park & Guan, 2006); however, this study focuses on the more macro collective face concept which while discussed by Chinese scholars has only been minimally investigated in an empirical study. Other constructs have been used to describe collective responses prompted by ingroup members’ transgression or success beyond Chinese contexts. This suggests that protecting or defending one’s own group is likely a pan-cultural phenomenon. For example, Shepherd, Spears, and Manstead (2013) discussed national shame felt by English students in response to ingroup members’ transgression (such as charging extra tuition for outgroup members). A second construct addressing the tension between individuals and their ingroup members is Cialdini et al.’s (1976) concept of “basking in reflected glory” (BIRG) in which people tend to associate themselves with successful ingroup members while cutting off reflected failure to dissociate themselves from unsuccessful ingroup members. An earlier study by Back (1951) found that when group members were primed by their group’s prestige, they risked little and acted cautiously so they would not endanger their group status. Mackie, Devos, and Smith (2000) and Smith and Mackie (2015) argued when people’s social identity becomes salient to interpret issues on a group basis, they will experience certain emotions shared within their 4 ingroup and such emotions have an effect on their actions towards the outgroup. Although these studies did not focus on collective face per se, they suggested that the concern with collective ingroup image in an intergroup communication context is likely a pan-cultural phenomenon. However, both Chinese and Western scholars who investigated collective group image did not provide a systematic conceptual framework about the concepts they explored. For Chinese scholars, their exploration about collective image was either based on personal observations (e.g., Hu, 1944) and they provided little quantitative evidence for their claims (e.g., Zhai, 2011). Western scholars (e.g., Back, 1951; Cialdini et al., 1976), on the other hand, overlooked people’s concern with the collective image of their own group perceived by the outgroup. Hence, Zhu (2014) proposed a systematic, hypothesis-driven, testable concept--- collective face---as a new approach to explore the collective image of a group. Collective face is defined as one’s concern with a collective positive image of one’s salient ingroup in the eyes of outgroups and related communicative strategies that this person adopts on behalf of the ingroup to ensure that the positive image will be maintained in the perception of outgroup members. People with high collective face concerns are worried about how members of outgroups may judge their group’s competence based on their own performance especially in response to a threat. Hu (1944) and Hail (2015) mentioned that Chinese were sensitive to their country’s face when meeting with foreigners. In a previous study (Zhu & Bresnahan, 2018), American and Chinese participants were asked to report their collective face concerns in response to an imagined face-threatening event, and the results demonstrated that while both groups revealed concern for collective face, Chinese had higher collective face concerns compared with Americans. These results were obtained from a quasi-experimental scenario study. The current 5 study tests whether collective face can be elicited using a controlled experiment. Given this earlier body of research, the following prediction about collective face is tested. H1: Participants’ cultural identities are likely to have an effect on collective face concerns in which Chinese participants are likely to report higher collective face concerns compared with American participants. Additional research is needed to investigate how ingroup versus outgroup status of a critic, who criticizes people’s ingroup members, influences their collective face concerns. Chinese and Americans are likely to have different responses depending on the ingroup-outgroup status of the person who criticizes the participant’s country. Zhai (2011) demonstrated several examples showing that failed Chinese Olympic athletes faced harsh criticism from other Chinese since they were considered to have lost face for China and for all Chinese people in front of the outgroup. For Chinese, an ingroup member who criticizes China in the eyes of the outgroup is similar to a failed Olympic athlete since both diminish face for their salient ingroup in the eyes of outgroup members. Hwang (2012) also argued that one’s success or failure is shared with relational others due to Chinese Confucian relationalism. For Chinese, criticism delivered by an ingroup member will result in higher collective face concern compared with criticism from an outgroup member because the outgroup member is likely to be perceived as having little association and intimacy with the salient Chinese community. As a result, it is reasonable to test the following prediction. H2: The ingroup-outgroup membership of the critic is likely to have an effect on collective face concerns for Chinese. Chinese participants in the condition where an ingroup member criticizes their country are likely to report higher collective face concerns compared with the condition where criticism is elicited from an outgroup member. 6 For members of Western individualist cultures, although only a few studies investigated face and country image, numerous studies demonstrate that ingroup transgression influences people’s shame and communicative response (e.g., Lyer, Schmader, & Lickel, 2007; Shepherd et al., 2013). For example, Lyer et al. (2007) investigated American and British’s response to their countries’ occupation of Iraq and found that messages that elicited high threat to their countries’ images increased feeling of shame which also predicted the intention to advocate for withdrawal from Iraq. However, ingroup transgression in the act of occupying another sovereign country derives from perception of ethical violations rather than from a challenge to group competence as threat to collective face in the current study. Second, ingroup transgressions in previous studies (e.g., occupation of another country) were induced by a government agency while they were beyond an individual ingroup (or outgroup) member’s control examined in the current study. Third, face threat is not same as the shame brought by ingroup transgression. People could feel ashamed privately without any audience being present, whereas face loss is often associated with the presence of others (Ho, 1976). The effect of the ingroup-outgroup membership of the critic on collective face is not clear for American participants. Therefore, the following research question is asked. RQ1: How does the ingroup-outgroup membership of the critic influence Americans’ collective face concerns? Previously, the author explored the interaction effect between cultural identities and ingroup-outgroup membership on collective face concerns in his preliminary paper; however, results lacked magnitude likely due to low experimental realism in the vignette-based study. The current study conducted an experiment to re-examine this question with several modifications of 7 the previous design by inducing a threat to participants’ collective face in a controlled laboratory setting. The experiment investigates this research question. RQ2: Is there any interaction effect between cultural identities and the ingroup-outgroup membership of the critic on collective face concerns in general? The next two hypotheses examine Chinese and Americans’ liking toward ingroup and outgroup members who offer criticism. If, as predicted in Hypothesis 2, Chinese participants experience higher collective face concerns in the condition where an ingroup member criticizes their country, they will show resentment for this ingroup member. Zhai (2011) discussed several examples of Chinese showing strong disapproval for ingroup members who failed to appear competent in front of foreigners (such as Chinese resentment and threats to Chinese athletes who performed poorly in the Olympics). The incompetent ingroup member threatens the positive image of the whole community. Being critical of your country in public in front of outgroup members is likely to make the Chinese critic himself or herself incompetent in terms of Chinese values for Confucian relationship, ingroup harmony, and face maintenance discussed earlier (see Hwang, 2012, for discussions on Chinese values). Chinese international students show their national pride and loyalty to their nation-state while discussing topics related to China with Americans (Hail, 2015). In contrast, an ingroup critic exposes the image of China under criticism and questions in front of outgroups. A traitor who betrayed his or her country violates the expectation that ingroup members should stay together. Likewise, an ingroup member who criticizes his or her own country in front of outgroups violates the whole community’s expectations that ingroup members should protect collective face for themselves and their national image. 8 The unacceptable “traitor” behavior of one person threatens all those who affiliate with this person whether directly or by virtue of their group identity, and the “betrayal” occurs because other people put their faith, hopes, and dreams on this person who failed to produce the desired outcome or to uphold the social norm. This failure derogates the public image of all and is outside of other ingroup members’ control. Everybody affiliated with this person, whether directly or indirectly, suffers collective face loss. In order to compensate for the face threat elicited by this traitor, collective face concerns are warranted here for Chinese participants to present their patriotic status and to respond to betrayal in an aggressive manner via showing strong resentment or disapproval to this traitor like what happened to unsuccessful Chinese athletes. Such resentment or disapproval will lead to low liking towards the ingroup member who disappoints the ingroup. As for an outgroup member who criticizes China, he or she is not as likely to induce strong collective face concerns in comparison to an ingroup member. As a consequence, he or she will receive less negative evaluation from Chinese compared with the ingroup member. However, in both conditions where criticism appears, Chinese participants’ high collective face concerns are likely to result in disapproval for such threat to their country image. The more collective face a Chinese has, the more likely he or she will show resentment towards this person who threatens his or her country’s image regardless of this critic’s membership. Therefore, a hypothesis is proposed. H3: There is a collective-face effect on liking towards the critic for Chinese: a) Chinese are likely to show lower liking towards the ingroup critic compared with the outgroup critic and b) Collective face concerns are negatively associated with Chinese participants’ liking towards the person who criticizes China in both the ingroup-critic and outgroup-critic conditions. 9 As noted by social identity theory, people follow ingroup favoritism so they can present themselves positively compared with outgroups (see Hogg et al., 1995 for a detailed discussion). Crocker, Thompson, McGraw, and Ingerman (1987) found evidence about ingroup favoritism among American campus sorority members where the ingroup was evaluated less negatively and more positively than members of outgroups. However, their study did not examine whether ingroup favoritism persists when the ingroup member acts like a “traitor” who attacks his or her own ingroup image. In addition, evidence shows that Americans also have collective face concerns (Zhu & Bresnahan, 2018). Since collective face concerns are predicted to cause resentment or disapproval for Chinese as discussed in the previous section, it is questionable whether the same effect applies to Americans. Due to the limited evidence, a research question is asked to explore Americans’ liking towards the critic across the ingroup and outgroup conditions and if collective face is related to Americans’ liking towards the critic. RQ3: a) Are there differences between the ingroup-critic and outgroup-critic conditions regarding Americans’ liking towards the critic? b) What are the relationships between Americans’ collective face concerns and their liking towards the critic? Participants’ positive and negative affect induced by the study is also examined. An earlier study suggested a connection between Chinese collective face concerns and their negative affect (discomfort feelings) when they read a scenario describing another Chinese international student suffering from face threat (Zhu & Bresnahan, 2018). A Chinese in the condition where an ingroup member criticizes China will experience high collective face concerns (see Hypothesis 2). At that moment, this Chinese participant’s collective face concern turns into resentment and disapproval towards this ingroup member since this person is considered as a traitor to his or her country (see Hypothesis 3), which results in less positive affect and more 10 negative affect for Chinese participants. As for the Chinese outgroup-critic condition, even though Chinese participants will not experience such high collective face concerns compared with ingroup criticism, the presence of the outgroup critic still makes Chinese worried about whether their behaviors may influence outgroups’ perception of their country (see Hail, 2015; Hu, 1994). This may lead to similar results like what occurs in the ingroup-critic condition. Thus, the following question is investigated. H4: Collective face concerns are likely to be associated with less positive affect and more negative affect for Chinese in both the ingroup-critic and outgroup-critic conditions. It is unclear about Americans’ liking towards the ingroup and outgroup critic (see Research Question 3) not to mention Americans’ affect. Hence, a research question is asked. RQ4: What are the relationships between collective face concerns and American participants’ positive and negative affect? In addition to participants’ liking towards the critic and their affect aroused by the study, it is unknown how they respond to other noncritical outgroup witnesses in this context. Based on the definition of collective face discussed in the beginning of this manuscript, the presence of the outgroup member is necessary for a collective face response to occur (e.g., Hu, 1944; Zhu, 2014). Numerous theories discuss reaction to and perception of outgroup members. For example, people may experience ingroup favoritism and make downward comparison toward the outgroup (Crocker et al., 1987; Hogg et al., 1995) or derogate outgroup members for people’s own self-image maintenance (Fein & Spencer, 1997), which contributes to stereotypes and prejudice. Outgroups are seen as a threat to self-concept in general in these theories. However, how does an individual deal with the mere presence of outgroup members if this person has nothing to do with acts that threaten ingroup members’ collective face? People might become 11 more likely to defend their collective image to restore the positive evaluation of their group from these noncritical witnesses or they may not be bothered at all by the presence of the outgroup witness simply because the outgroup plays a less salient role in their life. A research question is asked to investigate behavioral response to outgroup members who are present but do not offer criticism. RQ5: What is participants’ liking towards the noncritical outgroup witness (who did not induce any threat to participants’ collective face)? METHOD Participants and Procedure A total of 215 participants (100 American students and 115 Chinese international students) were recruited at a Midwestern university. American participants were recruited from several Communication classes while Chinese were recruited in several buildings where many Chinese students gathered. Each participant received a fifteen-dollar Amazon gift card in compensation for participation in this study. The study was a 2 (cultural identities: Americans or Chinese) by 3 (the group membership: the critic was either an ingroup member, an outgroup member, or had an unspecified identity to participants) between-subjects experimental design. This design resulted in four experimental conditions: 1) American participants with an American critic (n = 35); 2) American participants with a Chinese critic (n = 41); 3) Chinese participants with a Chinese critic (n = 34); and 4) Chinese participants with an American critic (n = 50). The first and third conditions were ingroup-critic conditions and the second and the fourth were outgroup-critic conditions. The experimental conditions also included a non-critical witness who was identified as an international student from Sweden to create an intercultural communication context. There were two control conditions where Americans (n = 24) or Chinese (n = 31) 12 interacted with the critic and the witness in which both of them did not disclose their national identities. In all cases, a confederate following a script as part of the experiment played the critic and the witness. Participants were asked to discuss two topics, environmental issues and drug use in the Olympics. These two topics were selected because both topics are related to American and Chinese collective face on a global level. Both Americans and Chinese are concerned with global warming (Koehn, 2008) and both American and Chinese athletes were investigated for doping during the Olympics (Gleaves, 2015; Yang & Leung, 2008). The sequences of topics were counterbalanced to eliminate any ordering effect. Including two topics aimed to broaden the scope of the study and to strengthen collective face response since the critic threatened participants’ collective face repeatedly in both discussions. Participants were told that the goal of this study was to examine how important social issues were discussed in the online chatroom. When the participant arrived in the research facility, an assistant asked about participants’ national identity and then decided which one of three membership conditions (the critic either identified as an ingroup member, an outgroup member, or someone with an unspecified identity to participants) the participant was assigned to with the help of a randomizer. The participant was guided to a small laboratory room, and given basic instructions about how to use the laptop to communicate online with other participants. The participant found an online chat room open on the laptop. The chat software used in this study was Skype in which participants can see texts typed during the conversation. The video camera function in Skype was not used during the chat and the participant’s profile image in Skype was set to the default to minimize the effect of physical appearance and other visual cues. At the beginning of the study, participants signed into the chat room using a fictional ID 13 assigned to them. Once the participant signed in, he or she saw that three other people were already in this chat room. For the purposes of this experiment, the research confederate played all three other roles based on the pre-selected script via three Skype accounts. These roles included the research assistant who served as the monitor of the study, the critic who was going to offer negative comments criticizing the participant’s country, and one outgroup witness who made noncritical comments. The research confederate entered all the comments in the chatroom using a given script based on participants’ conditions (see sample scripts in Appendix A) except for comments entered by the research participants. The critic disclosed his or her national background at the beginning of the study to identify as either an American or Chinese depending on which experimental condition was assigned. During the course of the online conversation, the “critic” made statements threatening participants’ collective face via criticizing the participant’s country for environmental issues and doping in Olympics The noncritical outgroup member was identified as a student coming from Sweden (a scripted role also played by the confederate). This noncritical role was included in this study to induce a high-level collective face response and make each condition an intercultural communication context. If a noncritical third party outgroup member witnesses what could be seen as a threat to someone else’s collective face, participants may be likely to experience stronger collective face response since the incompetence of one’s country is presented in front of an audience from different countries instead of the U.S. and China alone. Second, based on the conceptual definition of collective face (Zhu, 2014; Zhu & Bresnahan, 2018), the presence of the outgroup member is necessary for collective face to occur. If the noncritical outgroup member comes from either the U.S. or China, then in one of the ingroup-critic conditions, all three 14 members (the participant, the critic, and the noncritical witness) share the same national identity, which makes it hard to elicit any collective face. Third, if this noncritical member is either American or Chinese, the research participants in the ingroup conditions may experience low collective face response since the responsibility to carry on collective face is diffused among multiple ingroup members. Fourth, this outgroup member has to come from a country like Sweden with which participants are not very familiar. Otherwise, a country which shares too many similar features with either the U.S. or China may interfere with the outgroup status or participants will hold certain stereotypes about this outgroup member. For example, Americans may not consider a British witness an outgroup member at all due to the strong similarities between people in the U.S. and Great Britain. In the control conditions, participants still discussed two topics related to environmental issues and the Olympics and the sequences of the discussion topics were counterbalanced as well. However, in the control conditions, neither the critic nor the noncritical witness disclosed their national identities at the beginning of the conversation and the critic did not pose a specific face threat in either of the two topics during the conversation. The critic just simply addressed that some countries were responsible for environmental issues and doping in the Olympics while no specific comments targeting any countries were provided. The control conditions aimed to explore participants’ default response while eliminating all possible inductions of collective face such as disclosing national identities and threatening another person’s country via negative comments. After the online discussion was concluded, participants were directed to a Qualtrics survey on the laptop. Their collective face concerns, liking towards the critic and the witness, positive and negative affect for the study, and demographic information were measured in the 15 survey. They were asked to leave their MSU email addresses in a separate website to receive a $15 Amazon gift card after they completed the survey so their identities would not be linked with their responses. At the end of the study, participants were debriefed. They were also asked not to disclose the content of the study to anyone else. The Amazon gift card was sent to the participant via their official MSU emails within one week after their completion of the study. Measures Seven-point Likert scales were used for all measures and the items for all measures are listed in Appendix B. Indicators for goodness of measurement model fit were assessed through confirmatory factor analyses via SPSS AMOS for each measure to ensure uni-dimensionality. Reliability analysis in SPSS was also conducted for each measure. The indices for measures are reported in Table 1. [Table 1 here] Collective face concerns. This five-item scale was adopted from Zhu and Bresnahan (2018). The scale measured a unidimensional construct. The items included: 1) My image is closely related to my country’s image; 2) I worry that other people might think badly of me if someone who shares the same national identity with me did something wrong; 3) My performance reflects the competence of those who share the same national identity with me; 4) Foreigners may evaluate my country based on my performance; and 5) I should try my best to perform better so foreigners would not look down on my country and my people. Liking towards the critic. This measure included five items. Two items were modified from Back (1951) and other items were created by the author. The scale measured a unidimensional construct. Sample items included: 1) I would enjoy talking to this person; 2) I 16 would discuss important personal problems with this person; and 3) This person looks very friendly. Liking towards the noncritical witness. This measure included five items. Two items were modified from Back (1951) and the author created other items. The scale measured a unidimensional construct. Sample items included: 1) I would like to see this person around campus sometimes; 2) I would like to make a friend with this person; and 3) Talking with this person was a good experience. Positive and negative affect towards the study. Positive affect was measured as participants’ liking towards the study itself by a four-item scale. One item was modified from Festinger and Carlsmith (1959) and other items were created by the author. The scale measured a unidimensional construct. Sample items included: 1) I enjoyed this study very much; 2) This study is meaningful for me.; and 3) This experience is really interesting for me. Negative affect was measured as participants’ discomfort feelings about the situation that occurred during the experiment by a four-item scale. Three items were modified from Elliot and Devine (1994) and one item was created by Zhu and Bresnahan (2018). The scale measured a unidimensional construct. Sample items included: 1) I felt uncomfortable during my discussion with other participants; 2) I felt uneasy during my discussion with other participants; and 3) I felt confused during my discussion with other participants. RESULTS Descriptive statistics by condition are reported in Table 2. Hypothesis 1 examined whether there was a main effect of cultures on collective face. A two-way ANCOVA was conducted to test this hypothesis. Gender was controlled as a categorical covariate in this test and all following ANCOVAs (please see the limitation section for further discussion about 17 including gender as a categorical covariate). The results showed a significant main effect of cultures on collective face, F (1, 208) = 40.465, p < .001, η2 = .159. Chinese (M = 4.910, SD = 1.181, adj. M = 4.930, SE = .112, n = 115) reported higher collective face compared with Americans (M = 3.886, SD = 1.150, adj. M = 3.873, SE = .121, n = 100). Gender was related to collective face, F (1, 208) = 4.385, p = .037, η2 = .017. The data were consistent with Hypothesis 1. [Table 2 here] Hypothesis 2 examined the effect for the ingroup-outgroup membership of the critic on collective face for Chinese. A one-way ANCOVA was conducted just for Chinese participants to test this hypothesis and gender was controlled as a categorical covariate in this analysis. The effect of the membership was insignificant, F (2, 111) = 0.556, p = .575 for Chinese. Gender had an effect on collective face among Chinese, F (1, 111) = 7.015, p = .009, η2 = .059. Hence, the data were not consistent with Hypothesis 2 predicting the effect of the ingroup-outgroup membership on collective face for Chinese. As for Research Question 1 exploring whether the ingroup-outgroup membership of the critic influences American collective face, a one-way ANCOVA was conducted just for American participants. Gender was controlled as a categorical covariate in this analysis. The data suggested there was no significant effect of the membership, F (2, 96) = 0.722, p = .488. In addition, gender was not related to collective face for Americans, F (1, 96) = .004, p = .949. The effect of the ingroup-outgroup membership on collective face among Americans were not observed in the data. In order to further examine Hypothesis 2 and Research Question 1, the same two-way ANCOVA which was used for testing Hypothesis 1 was conducted to test the effect of ingroup- 18 outgroup membership on collective face for all participants showing that the result was still not significant, F (2, 208) = 0.211, p = .810. Research Question 2 exploring the interaction effect of cultural identities and the group membership on collective face in the total sample was tested by the same two-way ANCOVA above showing the result was not significant, F (2, 208) = 0.929, p = .397. The interaction effect between cultural identities and membership was not significant. Hypothesis 3a predicted that Chinese are likely to show less liking for the ingroup critic than the outgroup critic. A one-way ANOVA for Chinese participants with gender as a categorical covariate revealed the effect of the membership was significant on liking, F (2, 111) = 5.947, p = .004, η2 = .092. A further post hoc test with a Bonferroni correction showed a significant mean difference of 1.052 between the outgroup-critic condition and the control condition for Chinese (SE = .305, p = .002). Chinese participants in the outgroup-critic condition (M = 3.396, SD = 1.441, adj. M = 3.359, SE = .189, n = 50) reported less liking towards the critic compared with the Chinese control condition (M = 4.374, SD = 1.220, adj. M = 4.411, SE = .239, n = 31) while the ingroup-critic condition for Chinese (M = 3.718, SD = 1.368, adj. M = 3.738, SE = .228, n = 34) did not differ from either of other two Chinese conditions. Gender was related to liking towards the critic as well for Chinese participants, F (1, 111) = 6.794, p = .010, η2 = .052. Therefore, the data were not consistent with Hypothesis 3a predicting a difference in liking towards the critic between the ingroup and outgroup conditions among Chinese. As for Research Question 3a investigating the effect of the ingroup-outgroup membership on liking towards the critic among Americans, a one-way ANOVA for American participants with gender as a categorical covariate revealed such effects were not significant, F (2, 96) = 0.440, p = .645. Gender was not related to liking towards the critic for American participants, F 19 (1, 96) = .331, p = .566. Americans’ liking towards the critic did not differ across the ingroup and outgroup conditions. In order to further explore Hypothesis 3a and Research Question 3a about liking towards the critic, a two-way ANCOVA with gender as a categorical covariate was conducted for all participants. The results suggested a main effect of cultures on liking towards the critic, F (1, 208) = 7.038, p = .009, η2 = .031. Americans in general (M = 4.246, SD = 1.248, adj. M = 4.298, SE = .136, n = 100) reported higher liking towards the critic compared with Chinese in general (M = 3.755, SD = 1.410, adj, M = 3.802, SE = .126, n = 115). This ANCOVA also showed a significant main effect of the ingroup-outgroup membership on liking towards the critic for all participants, F (2, 208) = 3.486, p = .032, η2 = .030. A further post hoc test with a Bonferroni correction showed a significant mean difference of 0.576 between the outgroup-critic conditions and the control conditions (SE = .226, p = .034). Participants in the outgroup-critic conditions in general (M = 3.796, SD = 1.396, adj. M = 3.828, SE = .138, n = 91) reported less liking towards the critic compared with the control conditions in general (M = 4.386, SD = 1.139, adj. M = 4.404, SE = .178, n = 55) while the ingroup-critic conditions in general (M = 3.910, SD = 1.412, adj. M = 3.917, SE = .158, n = 69) did not differ from either of other two types of group membership conditions regardless of cultures. Gender was not related to liking towards the critic in general, F (1, 208) = 2.947, p = .088. The interaction between cultures and the ingroup-outgroup membership on liking towards the critic was not significant for all participants, F (2, 208) = 2.033, p = .133. Hypothesis 3b predicted a negative association between collective face and liking towards the critic among Chinese in both the ingroup-critic and outgroup-critic conditions. This hypothesis was tested using correlational analyses. The data were split based on the six 20 conditions and the correlation between collective face and liking towards the critic was assessed in each condition. The results suggested such correlation between collective face and liking towards the critic was only significant in the Chinese outgroup-critic condition, r (48) = -.343, p = .015, but not in the Chinese ingroup-critic condition, r (32) = .166, p =.349, nor in the Chinese control condition, r (29) = .168, p = .365. Hence, the data were only partially consistent with H3b. There was a significant negative association between collective face and liking towards the critic in the Chinese outgroup condition, but not in the Chinese ingroup condition. Research Question 3b exploring the relationship between collective face and liking towards the critic among Americans was tested by correlational analysis. The results showed that this relationship in all three American conditions was not significant: r (33) = .042, p = .810 for the American ingroup-critic condition; r (39) = .040, p = .806 for the American outgroup-critic condition; and r (22) =.083, p = .699 for the American control condition. It is interesting to observe, among all six conditions discussed above, the relationship between collective face and liking towards the critic was only significant in the Chinese outgroup-critic condition which suggested the presence of the outgroup critic made Chinese participants’ collective face more salient. In response to criticism from an outgroup member, the higher collective face Chinese have, the lower liking they felt towards this outgroup critic. Hypothesis 4 predicted a relationship between collective face and affect among the Chinese ingroup-critic and outgroup-critic conditions. The data were first split by conditions and then were assessed with correlational analyses. Results showed that for Chinese participants, collective face was positively correlated with discomfort feelings (negative affect) in both the ingroup-critic, r (32) = .436, p = .010, and the outgroup-critic conditions, r (48) = .353, p = .012. However, for Chinese, collective face was not correlated with liking towards the study (positive 21 affect) in either the ingroup-critic condition, r (32) = .261, p = .136, or the outgroup-critic condition, r (48) = -.011, p = .938. Therefore, the data were only partially consistent with Hypothesis 4. Research Question 4 exploring the relationship between collective face and affect among Americans was assessed using correlational analyses. Americans’ collective face was positively correlated with liking towards the study in the ingroup-critic condition, r (33) = .425, p = .011, but not in the outgroup-critic condition, r (39) = .236, p = .138. In contrast, Americans’ collective face was positively correlated with their discomfort feelings in the outgroup-critic condition, r (39) = .354, p = .023, but not in the ingroup-critic condition, r (33) = .014, p = .936. These results suggested American participants’ collective face was associated with positive and negative affect differently depending on the critic’s membership. Research Question 5 investigating participants’ liking towards the noncritical outgroup member was assessed using both two-way ANCOVA with gender as a categorical covariate and correlational analyses. Neither national cultures nor the ingroup-outgroup membership yielded a significant main effect, F (1, 208) = 2.772, p = .097, and F (2, 208) = 1.317, p = .270 respectively. The interaction effect was not significant either, F (2, 208) = .353, p = .703. After splitting the data based on conditions, correlational analyses yielded an interesting pattern. American collective face was positively correlated with liking towards the noncritical outgroup member in both the ingroup-critic and outgroup-critic conditions, r (33) = .400, p = .017, and r (39) = .404, p = .009 respectively. Such correlation was not significant in each of the other four conditions (the three Chinese conditions plus the American control condition), with Pearson’s correlation coefficients ranging from -.243 to .271 and p-values ranging from .121 to .707. 22 Other Findings Based on the results from Hypothesis 4 and Research Question 4, differences in the correlations between collective face and affect (liking towards the study and discomfort feelings) were observed between American and Chinese participants in the ingroup-critic and outgroup- critic conditions. Further ANCOVAs with gender as a categorical covariate were conducted to examine possible main effects and interaction effects of cultural identities and the ingroup- outgroup membership on liking towards the study and discomfort in the four experimental conditions where collective face threats were elicited to target participants’ country. The control conditions did not induce specific collective face threat to participants’ country so the participants were not expected to experience any affect aroused by their collective face concerns. Therefore, the control conditions were excluded from the analysis exploring affect here. As for liking for the study, both main effects were not significant, F (1, 155) = .299, p = .585 for cultures, and F (1, 155) = .516, p = .474 for the ingroup-outgroup membership. The interaction effect on liking for the study was not significant, F (1, 155) = .307, p = .580. Gender was not related to liking towards the study, F (1, 155) = .001, p = .980. As for discomfort feelings, there was a significant cultural effect, F (1, 155) = 19.211, p < .001, η2 = .105. Chinese participants in both the ingroup-critic and outgroup-critic conditions together experienced more discomfort feelings (M = 3.378, SD = 1.396, adj. M = 3.387, SE = .144, n = 84) than American participants in these two conditions together (M = 2.523, SD = 1.210, adj. M = 2.463, SE = .150, n = 76). Gender was related to discomfort feelings, F (1, 155) = 6.511, p = .012, η2 = .035. However, the main effect of the ingroup-outgroup membership and the interaction effect on discomfort feelings were not significant, F (1, 155) = 1.091, p = .298, and F (1, 155) = 1.891, p = .171 respectively. 23 The cultural difference between Americans and Chinese in discomfort feelings observed above may result from the cultural difference in collective face between participants who identify with either of these two cultures (see Zhu & Bresnahan, 2018 for details). Hence, an OLS regression was conducted to examine if collective face had any main effect or interacted with conditions to influence discomfort feelings. For the same reason discussed above, the control conditions were not included in this analysis due to the lack of specific collective face threat targeting on participants’ country. The four experimental conditions were dummy coded in which the American ingroup-critic condition was treated as a reference group. Each of other three conditions was considered as a unique comparison group. The three dummy coded comparison groups, centered collective face (calculated excluding the two control conditions), and three interaction terms between centered collective face and each of dummy coded comparison groups were entered into the regression as predictors and feelings of discomfort was the outcome variable. The whole model showed goodness of fit, F (7, 152) = 5.929, p < .001, R2 = .214, adj. R2 = .178. The results yielded a significant main effect in the Chinese outgroup-critic condition, B = .786, β = .266, t = 2.648, p = .009. The mean of discomfort feelings among Chinese participants in the outgroup-critic condition was a .786 unit higher than the mean of discomfort feelings among American participants in the ingroup-critic condition when they both have an average collective face among these four experimental conditions of 4.43 (n = 160). Centered collective face itself had little impact on discomfort in the American ingroup-critic condition, B = .015, β = .014, t = 0.080, p = .936. However, the coefficient of the interaction term between the dummy coded Chinese ingroup-critic condition and centered collective face was significant, B = .588, β = .219, t = 2.063, p = .041. There was a difference in the slope predicting discomfort feelings by collective face between the Chinese ingroup condition and the 24 reference group. This suggested, as for per unit of an increase in collective face, Chinese who were exposed to an ingroup member’s collective face threat would feel a 0.588 more unit of discomfort feelings compared with Americans who were exposed to an ingroup member’s collective face threat (please see Table 3 for this regression output). [Table 3 here] DISCUSSION Cultural Differences and Traitor Effects Cultural differences between American and Chinese participants were observed for collective face, liking towards the critic, and discomfort feelings. Chinese reported higher collective face concerns compared with Americans, consistent with earlier findings (Zhu & Bresnahan, 2018). In addition, compared with Americans, Chinese reported lower liking of the critic. Among the four experimental conditions where the critic elicited specific negative comments targeting participants’ countries, Chinese also felt more discomfort compared with discomfort experienced by Americans. These results suggest that Chinese are more sensitive to their collective positive image and rate the critic who elicits collective face threat less favorably in intercultural communication contexts compared with Americans. Future research may explore how this finding can be applied to political, educational, and business settings. A question that remains is what will happen if the critic de-collectivizes such criticism to focus instead on personal responsibility for certain individuals rather than the whole country’s image to see how Chinese and Americans respond to such individualized criticism. Regarding the traitor effect which predicted that Chinese would be likely to show lower liking towards the ingroup critic than the outgroup critic, no significant difference was found between these two conditions. However, collective face was significantly correlated with liking 25 towards the critic in the Chinese outgroup-critic condition in a negative direction while this correlation remained insignificant in other conditions. One explanation for this result is that the relationship between collective face and dislike for the critic was only primed by the presence of the outgroup critic for Chinese. However, this possibility needs further investigation. Another interesting phenomenon was the correlation between collective face and negative affect (discomfort feelings) among the four experimental conditions where participants’ country image was specifically threatened. Such positive correlation was significant for all experimental conditions except the American ingroup-critic condition. In contrast, among the same four experimental conditions, the correlation between collective face and liking towards the study was only significant for the American ingroup condition, but not for the other three conditions. It is possible that Americans did not consider the ingroup critic’s comments as a collective face threat at all so they did not feel bothered, which resulted in little association between collective face and discomfort. Americans from an individualistic background tend to use dominating and assertive strategies to address conflicts (see Ohbuchi, Fukushima & Tedeschi, 1999; Ting-Toomey et al., 1991) so an ingroup member’s criticism was considered to be less threatening. Americans may feel more relaxed in the ingroup condition where they only need to address criticism from an ingroup member, which primes the relationship between collective face and liking towards the study. Chinese participants’ discomfort feelings in the ingroup-critic condition were more sensitive to their collective face compared with Americans in the same condition. In this condition addressing an ingroup member’s criticism, the higher collective face a Chinese has, the higher discomfort feelings he or she suffers. In contrast, Americans’ feelings of discomfort were 26 unlikely determined by their collective face even while encountering an ingroup critic criticizing the U.S. This may result from the traitor effect among Chinese described in the earlier section where an ingroup member who is supposed to defend his or her own country actually derogates its image. Results suggest that compared with American participants, Chinese participants are more likely to experience discomfort from collective face primed by the traitor effect. It is also likely that Chinese high collective face makes Chinese have certain stable expectations for an ingroup member’s behavior towards preserving his or her country’s image while any violation like an ingroup member expressing collective face threat can make this ingroup critic a traitor. There is likely a normative expectation that a Chinese person should not criticize China in public in the U.S., especially when outgroup members are present. Exposing potentially embarrassing details, regardless of whether they are factual or not, is clearly seen as disloyal by Chinese in this study. In contrast, American relatively low collective face makes it difficult for them to have a similar expectation. Therefore, even a violation in which an ingroup member threatened the image of the U.S. did not surprise Americans. This explains why Americans’ discomfort feelings were not as susceptible to collective face compared with Chinese in the ingroup-critic conditions. The traitor effect in the Chinese ingroup-critic condition made Chinese feel embarrassed and worried. Chinese felt disappointed and confused about why an ingroup member was turned into a traitor violating his or her country image. Implications The current study investigated communication within and between two cultural groups in an intercultural communication context where their country’s image was criticized. This study replicated a previous study (Zhu & Bresnahan, 2018) exploring how collective face interacts 27 with cultures and ingroup-outgroup membership in an intercultural face-threatening context and provided more empirical evidence for cultural differences in collective face concerns between Chinese and Americans in a laboratory experimental design. Zhu and Bresnahan’s (2018) study showed while witnessing an ingroup member’ embarrassment, Chinese participants’ discomfort feelings were more influenced by their collective face than Americans in the same context. The current study yielded similar results. While encountering an ingroup critic’s comments about China, Chinese participants’ discomfort feelings were more easily aroused by their collective face compared with Americans encountering an ingroup member’s criticism about the U.S. The presence of other ingroup members strengthens the relationship between collective face and negative affect for Chinese, but not for Americans. Mackie et al. (2000) found that group-based appraisal such as perceived societal support for one’s ingroup position was positively related to ingroup members’ anger towards the outgroup and tendency to take negative actions against the outgroup. Anger also mediated the relationship between such appraisal and offensive action tendencies (Mackie et al., 2000). In the current study, collective face, like group-based appraisal, is related to Chinese participants’ discomfort when an ingroup member criticizes China. Future studies may explore the conceptual similarity between group-based appraisal and collective face and investigate whether anger can also explain the traitor effect caused by collective face. In summary, the current study helps to further explore the nature of collective face and its affective outcome in complex intercultural communication contexts where both ingroup and outgroup statuses are examined. In addition, the current study suggested either ingroup favoritism or dissociation from ingroup members who violated group expectations is too simplistic to explain complexities in the intercultural communication context where people’s 28 collective face is under threat. The target and source of such face threat, the presence of the outgroup members, and people’s original collective face concerns, all should be taken into consideration to examine people’s attitudinal and affective responses in such contexts. Practical implications. Results of this study also yield practical implications for addressing intercultural conflict and communication between Americans and Chinese. As for a Chinese who wants to share with others his or her constructive criticism towards China in the intercultural context where both outgroup and ingroup members are present, this Chinese critic should focus on a specific individualized agent rather than their country to avoid collective face threat to other ingroup members who are present. A Chinese critic should consider the traitor effect and be aware that his or her criticism will turn other Chinese witnesses’ collective face into discomfort feelings, which may interfere with their ability to assess such criticism in an objective way. There is an old Chinese saying that home truth should not be shared with outsiders. A Chinese critic should find a balance between offering constructive opinions about China and maintaining collective face for himself or herself and all other Chinese audience in this context. If an outgroup member such as an American really wants to offer some constructive criticism towards China for a Chinese audience, they might do this in front of the target audience without any presence of other Chinese. Otherwise, both the target and the witnesses will experience negative feelings due to their collective face. The target audience will feel this outgroup critic intentionally puts themselves into a dilemma in which they are expected by the Chinese witnesses to defend the image of China. If not, the Chinese audience may risk themselves being perceived by their ingroup members as traitors who fail to protect their shared collective face. Outgroup critics may also consider offering messages to restore Chinese 29 collective face as well after their criticism. Otherwise, Chinese collective face will make Chinese consider such criticism, regardless of its constructive values, as a personal attack. In other words, outgroup critics should be careful about whether or not their messages are interpreted as a collective face threat by Chinese. Methodological implications. The current study also provided methodological implications for intercultural communication research. First, few intercultural communication studies have been conducted in a laboratory experimental setting with a relatively large international student population. The strategies and even the difficulties related to recruitment and research procedures can serve as guides for future research aimed at studying international students. In the beginning of the data collection, the researchers had some troubles in recruiting enough Chinese participants. Several strategies were used to increase participation. These included asking the Chinese student associations for help, using snowball sampling, asking the office of the registrar to email Chinese students, promoting the study via Chinese social media platforms with the help of the international student service office, asking the English Language Center to promote the study, and even recruiting Chinese students to distribute posters for the study. Among all of strategies used, recruiting other Chinese to promote the study worked better than other approaches. This may also result from a collective face phenomenon in which Chinese want to maintain a cooperative and harmonious group image in a U.S university. If a Chinese student refused to participate in a study organized by the Chinese research assistant, this student’s unwillingness to help other Chinese may receive unfavorable evaluations from their ingroup members. If Chinese students just receive an email promoting the study from their class instructors or some university agencies, their collective face will not be primed and they will have little motivation to participate in such a study. 30 The current study offered Amazon gift cards as an incentive for participation while most Chinese participants did not find it appealing enough for them to come to the laboratory. Future research may consider better incentives for recruiting Chinese participants since many Chinese may not consider a $15 gift card as a good motivator for them to physically come to a research facility. Chinese economic development enables many middle-class families to send Chinese students to the U.S. (Svoboda, 2015) so such a small amount of incentive is not comparable to the huge amount of tuition these Chinese students paid to the universities. Many Chinese students were unfamiliar with the lab experimental design so they refused to participate. It is also possible that Chinese were concerned with individual scrutiny, privacy, and disclosure of their personal attitudes required in an experiment. Even though anonymity and guarantees of privacy were made to participants, consistent with IRB policy, participants may have feared that anything they said in this study as international students could be linked back to them and come under the scrutiny of U.S. Homeland Security and Chinese authorities who control their visa to study in the U.S. In the future, researchers who are interested in an international student population should cooperate with the international student service office on campus to help international students be aware of the norms and cultures for participating in laboratory research at U.S. campuses. Training about research expectations and protection of private information should be included in international student orientations. This may help to increase research participation rates among international students. Additional work needs to be conducted on the traitor effect in cross-cultural contexts. Perhaps Chinese in this experiment showed more sensitivity to the traitor effect because they were international students in a foreign country and Americans were on their home ground. 31 Would American students studying in China be more alert to country criticism than they were in this experiment? Are Chinese domestic students more aware of foreigners’ criticism related to their country in China? In what contexts is the traitor effect most poignantly experienced? Future studies need to investigate the impact of the traitor effect on collective face arousal. Limitations First, in spite of using a randomizer for randomized assignment to conditions in this study, the current study had unequal sample sizes across conditions, which could affect ANOVA results. For instance, the small sample size in control conditions tended to yield a larger variance compared with variances in other conditions, which violates the assumption of homogeneity for ANOVA design. In response to this issue, Levene’s tests of homogeneity were conducted to check variances across the six conditions for all measured variables, and the results were not significant, Fs (5, 209) ranged from 0.578 to 1.501 with p-values ranging from .191 to .717. This suggested that the assumption of homogeneity was not violated and unbalanced sample sizes did not bias analyses. However, a future study should more closely monitor recruitment of relatively equal sample sizes across conditions. Second, the manipulation of the ingroup-outgroup membership of the critic was weak and all predictions addressing the differences between the ingroup and outgroup conditions were not supported by the data. The weak manipulation of ingroup-outgroup membership suggests, the online chatroom may have made it difficult for participants to recognize the critic’s national identity without any nonverbal cues. Although the critic identified as either an American or a Chinese at the beginning of the discussion, participants may have not attended to this detail or forgotten this earlier identification in the long conversation that followed. Repetition of national identity throughout the script such as statements like, “As an American or Chinese, what you just 32 said hurts me” might have reinforced the strength of group manipulation. Future studies may consider recruiting an authentic American or Chinese student as a confederate-critic to communicate with participants face-to-face to increase study realism and strengthen the manipulation of ingroup-outgroup membership. However, such procedures will bring more factors into consideration, such as the critic’s English competence, tone, physical attractiveness, and gender, which requires more time and money for training and may introduce confounds to the experiment. Third, the current study recruited a convenience sample using a college student population. The total sample showed more females participated in the study than males, especially for American participants who were recruited via several communication required classes while the department of communication in the university where the study was conducted has more female students enrolled than male students. Chinese participants were recruited in multiple methods so the gender ratio among Chinese was not as imbalanced as American participants. In the future, similar recruitment strategies should be applied to both cultural groups. However, the entire recruitment procedure for Chinese participants as mentioned before was both costly in terms of time, money, and reluctance shown for participation in an experiment. If similar recruitment strategies were applied to the American sample, that may have raised more logistic and economic concerns. Even gender was related to some of the variables in this study based on the results, it is too premature to make conclusions about the gender effect on collective face based on the current sample with an imbalanced gender distribution. In order to avoid possible bias due to this imbalanced gender ratio, gender was controlled as a categorical covariate in ANCOVAs in the current study to obtain more accurate results for the variables the author is interested in investigating. 33 Fourth, among American participants, about 40% participants identified as minority ethnic group members. In contrast, more than 85% of Chinese participants belonged to Han ethnic group, which is the majority ethnic group in China. Future studies should take such ethnicity within cultural groups into consideration to control its effects. However, such difference in ethnic compositions across Americans and Chinese may also reflect the demographic-ethnic characteristics of each cultural group. Fifth, all Americans were recruited before 2018 while the majority of Chinese participants were recruited after 2018. In 2018, the trade conflict between the U.S. and China started (Bown, 2019, published online). Due to the trade conflict between these two countries, people’s national pride and collective image may get primed more, which can also explain cultural differences observed in this study. However, the author questioned such priming effects since the experimental induction in the current study was focusing on different topics rather than international trade and business policies. Moreover, few Chinese participants mentioned the ongoing trade conflict during their conversation with the critic. The consistency between the current study and the previous study conducted by Zhu and Bresnahan (2018) also suggested such cultural differences were not the results of a political debate. On the other hand, future laboratory experimental studies exploring two cultural groups should limit data collection time to avoid the effect of possible confounding external influences. Sixth, participants in all conditions reported relatively low discomfort feelings which were below the midpoint of the seven-point scale. Even though significant cultural difference was observed between Americans and Chinese among these experimental conditions, the magnitude of the effect size was low. One possible explanation is social desirability. Participants may have been reluctant to report their negative feelings to avoid negative 34 evaluations from researchers. Reporting high negative feelings potentially makes participants seem powerless and confused, which is likely to threaten their positive image. Therefore, participants may be more likely to under-report their discomfort. The low discomfort score found in this study did not necessarily mean participants experienced little negative affect due to the critic’s collective face threat. The conversations in response to the criticism showed many participants felt their country image was stigmatized and they felt being offended. There was one female Chinese participant who even told the critic to stop criticizing China, otherwise she would report the critic. In response to this participant, we had both American and Chinese research assistants explain the true purpose of the study so no one in the study actually tried to derogate her country’s image. Future studies should consider using more unobtrusive measures, such as psychological tests like the Skin Conductance Level (see van Harreveld, Rutjens, Rotteveel, Nordgren, & Pligt, 2009), or response time tests (see Smith & Henry, 1996) to assess negative affect to avoid possible social desirability effects. Seventh, Chinese international students have heterogeneous experiences abroad (see Heng, 2019, published online) so the Chinese sample from a single Midwestern university may not be generalized to other Chinese international students in the U.S. Other factors, such as the duration of sojourn in the U.S., English-language competence, and academic performance may also influence their reactions to conversation with the critic. All American participants used their native language in the online chatroom while Chinese participants had to use a foreign language to communicate with the critic, which may also affect their response. A criticism delivered by a Chinese student critic to a Chinese student participant in English might be doubly facethreatening and embarrassing to collective face reinforcing the traitor effect. Future research should control these factors such as providing an online Chinese-English dictionary, measuring 35 Chinese students’ English proficiency, and giving Chinese students more time to think and type in English. Eighth, the Chinese international students’ unique international experience may prime their national identity and loyalty to their country (see Hail, 2015), which may also influence their collective face or may even become a confounding factor. The observed cultural differences in the current study may simply result from the differences between domestic and international students rather than Chinese or American cultural values. A future study should explore how American international students in China experience collective face responses while addressing negative comments about the U.S. to see if they behave in the similar patter as the Chinese international students did in this study. CONCLUSION The current study provides evidence about collective face and its affective outcome from two different cultural groups through a laboratory experimental design using an American and a Chinese international student population. Cultural differences in collective face, liking towards the critic, and discomfort feelings associated with collective face were observed. There was little research examining similar populations with laboratory experimental designs and the current study yielded interesting implications for future researchers who want to conduct a laboratory study with an international student population. Future collective face research should also explore the positive effect of collective face in promoting people’s national pride rather than just focusing on collective face threats and negative affect. Participants other than American domestic students or Chinese international students should be recruited in collective face research as well as American international students in China, Chinese domestic students, participants from other countries, and adult populations rather than a university student sample. 36 Collective face, as a newly developed concept, needs to be explored in multiple cultural contexts with more diverse populations, and using different research methods to demonstrate its theoretical and practical implications for the intercultural communication discipline and greater understanding of the psycho-social mechanisms underlying cross-cultural and intercultural communication. 37 APPENDICES 38 APPENDIX A Sample Scripts Assistant = Brian; Participant = Spirit; Critic = Will; Noncritical Witness = Courage (This noncritical witness was the Swedish outgroup member among the four experimental conditions but did not disclose his or her national identity in the two control conditions). All three characters except Spirit were played by a research confederate via three Skype accounts. All participants received a fictional ID of Spirit in which they believed the ID was assigned via a random drawing. Sample Script used for experimental conditions: Among the four experimental conditions, the underlined parts include fictional names and criticisms against participants’ country image. The bold parts include the manipulation of the membership of the critic via changing the critic’s national background information. The target of criticism and the national background information of Will and Courage vary based on participants’ national identities and three types of membership conditions. The following script was used for the American ingroup-critic condition. Beginning the conversation: Assistant: Hello, everyone. My name is Brian and I am working as an assistant to help you to complete this experiment. This study is investigating how online chat rooms influence people’s discussion about several issues. Do you have any questions so far? Critic: Nope. Witness: No problem. Participant: … Assistant: Let’s introduce yourselves first. Since this is an anonymous study, please do not disclose your actual name. But maybe you can let everyone know where you come from and what is your major or anything you’d like to share. Make sure that you also remember everyone’s ID in this chat room. Let’s begin with you, Courage. 39 Witness: Hi, everyone! I come from Sweden. I am majoring in computer engineering. I am an exchange student. My English is not good so let me know if my grammar is not understandable. Assistant: Good. Then you, Will. Critic: Hi. I was born in Michigan. This is my junior year. My major is communication. Assistant: How about you, Spirit. Participant: ... Assistant: Good. Are you ready so we can move on? Everyone (including P) response: Yes. Assistant: Ok. Let’s take one or two minutes to think about what is today’s most important environmental issue on a global basis. Are we good to go? Witness: Yes. Critic: Yes. Participant: … Assistant: Courage, what do you think? Witness: Air pollution, some countries built up so many factories. Release co2 or so2. That’s insane! Assistant: Ok, air pollution, could you specify that a little bit? Witness: lol… There are many regulations about releasing air pollution in many countries. But it is still a big concern globally. Air pollution influences our daily life, health, economy, politics, and many other things you know. Critic: I just learned from my Environmental Science class that US is responsible for air pollution and global warming but it makes no effort to solve that at all. This is terrible since US is incapable of addressing its environmental issues. That directly leads to global warming today. US increasing economic development sacrifices world environment! Witness: terrible… Participant: … (If participant did not say anything yet, the assistant will ask the participant directly about his or her opinion about other three commenters) If the participant starts to defend his or her country (e.g., “this is ridiculous about your comments to my country”; “stop attacking my country”). Critic: Come on! I am talking about air pollution. If the participant starts to question the critic’s source (e.g., “where did you get this information”, “sounds unrealistic”). Critic: I learned from an Environmental Science class. I forget its course number. If the participant starts to show some evidence to refute the critic (e.g., “My experience is not the same…”). Critic: Really? Not sure. If the participant starts to show agreement towards the critic (e.g., “You are right about that”). Critic: See, I am right. The witness will insert some emoticons and simple valence-free comments during the conversation between the participant and the critic (e.g., “lol”, “wow”, “really?”). (After a few round) Assistant: Good job. Anything else you want to share? Witness: No. 40 Critic: Nope. Participant: … Assistant: Ok, time out. Let’s move on to the next topic. Assistant: Ok. Here is a good topic. Olympics. What is the most controversial thing in Olympics. Let’s take one or two minutes to think about this topic. Assistant: Are we good to go? Critic: Yes. Witness: yeah. Participant: … Assistant: Any volunteer? Critic: Drug! Using unpermitted drug in games. You know like steroids. Unfair to Olympic spirit at all! Athletes just did that for medal sake. So stupid. I hate to mention that but US athletes had bad reputation for using drug in Olympics. Vanity I guess. Shame on them! Witness: I have no knowledge about that before. But it sounds too bad. Participant: … (If the participant did not say anything yet, the assistant will ask the participant directly about his or her opinion about other three commenters). If the participant starts to defend his or her country (e.g., “this is ridiculous about your comments to my country”; “stop attacking my country”). Critic: Come on! I am talking about Olympics. If the participant starts to question C’s source (e.g., “where did you get this information”, “sounds unrealistic”). Critic: I learned from some website, BBC I think. If the participant starts to show some evidence to refute the critic (e.g., “My experience is not the same…”). Critic: Really? Not sure. If the participant starts to show agreement towards the critic (e.g., “You are right about that”). Critic: See, I am right. The witness will insert some emoticons and simple valence-free comments during the conversation between the participant and the critic (e.g., “lol”, “wow”, “really?”) (After a few round) Assistant: Good job. Anything else you want to share? Critic: No. Witness: No. Participant: … Assistant: Ok, good job. Assistant: Let’s complete this online survey. The survey is about your participation in the study. Once you finish that, please open the door and talk with the assistant outside. (Sending the link) Sample Script used for control conditions: 41 In these two control conditions, the critic only offered general negative comments without targeting at a specific country and both the critic and the witness did not disclose their national backgrounds. The following script was used for both the American and Chinese control conditions. Beginning the conversation: Assistant (A): Hello, everyone. My name is Brian and I am working as an assistant to help you to complete this experiment. This study is investigating how online chat rooms influence people’s discussion about several issues. Do you have any questions so far? Critic (C): Nope. Witness (W): No problem. Participant (P): … Assistant: Let’s introduce yourselves first. Since this is an anonymous study, please do not disclose your actual name. But maybe you can let everyone know where you come from and what is your major or anything you’d like to share. Make sure that you also remember everyone’s ID in this chat room. Let’s begin with you, Courage. Witness: Hi, everyone! I am majoring in computer engineering. Assistant: Good. Then you, Will. Critic: Hi. This is my junior year. My major is communication. Assistant: How about you, Spirit. Participant: ... Assistant: Good. Are you ready so we can move on? Everyone (including P) response: Yes. Assistant: Ok. Let’s take one or two minutes to think about what is today’s most important environmental issue on a global basis. Assistant: Are we good to go? Witness: Yes. Critic: Yes. Participant: … Assistant: Courage, what do you think? Witness: Air pollution, some countries built up so many factories. Release co2 or so2. That’s insane! Assistant: Ok, air pollution, could you specify that a little bit? Witness: lol… There are many regulations about releasing air pollution in many countries. But it is still a big concern globally. Air pollution influences our daily life, health, economy, politics, and many other things you know. Critic: I just learned from my Environmental Science class that some countries are responsible for air pollution and global warming but they make no effort to solve that at all. This is terrible since they are incapable of addressing their environmental issues. That directly leads to global warming today. Their increasing economic development sacrifices world environment! 42 Witness: terrible… Participant: … (If participant did not say anything yet, the assistant will ask the participant directly about his or her opinion about other three commenters) If the participant starts to defend his or her country (e.g., “this is ridiculous about your comments to my country”; “stop attacking my country”). Critic: Come on! I am talking about air pollution. If the participant starts to question the critic’s source (e.g., “where did you get this information”, “sounds unrealistic”). Critic: I learned from an Environmental Science class. I forget its course number. If the participant starts to show some evidence to refute the critic (e.g., “My experience is not the same…”). Critic: Really? Not sure. If the participant starts to show agreement towards the critic (e.g., “You are right about that”). Critic: See, I am right. The witness will insert some emoticons and simple valence-free comments during the conversation between the participant and the critic (e.g., “lol”, “wow”, “really?”). (After a few round) Assistant: Good job. Anything else you want to share? Witness: No. Critic: Nope. Participant: … Assistant: Ok, time out. Let’s move on to the next topic. Assistant: Ok. Here is a good topic. Olympics. What is the most controversial thing in Olympics. Let’s take one or two minutes to think about this topic. Assistant: Are we good to go? Critic: Yes. Witness: yeah. Participant: … Assistant: Any volunteer? Critic: Drug! Using unpermitted drug in games. You know like steroids. Unfair to Olympic spirit at all! Athletes just did that for medal sake. So stupid. I hate to mention that but some countries had bad reputation for using drug in Olympics. Vanity I guess. Shame on them! Witness: I have no knowledge about that before. But it sounds too bad. Participant: … (If the participant did not say anything yet, the assistant will ask the participant directly about his or her opinion about other three commenters). If the participant starts to defend his or her country (e.g., “this is ridiculous about your comments to my country”; “stop attacking my country”). Critic: Come on! I am talking about Olympics. If the participant starts to question C’s source (e.g., “where did you get this information”, “sounds unrealistic”). Critic: I learned from some website, BBC I think. If the participant starts to show some evidence to refute the critic (e.g., “My experience is not the same…”). Critic: Really? Not sure. If the participant starts to show agreement towards the critic (e.g., “You are right about that”). Critic: See, I am right. 43 The witness will insert some emoticons and simple valence-free comments during the conversation between the participant and the critic (e.g., “lol”, “wow”, “really?”) (After a few round) Assistant: Good job. Anything else you want to share? Critic: No. Witness: No. Participant: … Assistant: Ok, good job. Assistant: Let’s complete this online survey. The survey is about your participation in the study. Once you finish that, please open the door and talk with the assistant outside. (Sending the link) 44 APPENDIX B Measures Collective face concerns (See Zhu & Bresnahan, 2018 for a detailed discussion of this measure). 1. My image is closely related to my country’s image. 2. I worry that other people might think badly of me if someone who shares the same national identity with me did something wrong. 3. My performance reflects the competence of those who share the same national identity with me. 4. Foreigners may evaluate my country based on my performance. 5. I should try my best to perform better so foreigners would not look down on my country and my people. Likings towards the critic (Items 1 and 2 are modified from Back, 1951) 1. I would enjoy talking to this person. 2. I would discuss important personal problems with this person. 3. This person looks very friendly. 4. I would like to make a friend with this person. 5. It is nice to meet this person. Liking towards the noncritical witness (Items 1 and 2 are modified from Back, 1951) 1. I would like to see this person around campus sometimes. 2. I would discuss important personal problems with this person. 3. This person looks very friendly. 4. I would like to make a friend with this person. 5. Talking with this person was a good experience. Likings towards the study itself (Item 1 is modified from Festinger and Carlsmith, 1959) 1. I enjoyed this study very much. 2. This study is meaningful for me. 3. This experience is really interesting for me. 4. I wish that there will be more similar experiments like this. Discomfort feelings (Items 1, 2, and 3 are modified from Elliot and Devine, 1994) 1. I would feel uncomfortable in this situation. 2. I would feel uneasy in this situation. 3. I would feel bothered in this situation. 4. I would feel confused in this situation. 45 Table 1: Measure Indices α GFI CFI RMR RMSEA SRMR χ2 df .065 .122 .084 .043 .066 .034 .042 .055 .025 .022 .037 .085 .027 .090 .047 .148 .162 .027 .089 .037 .018 .075 .000 .061 .089 .000 .000 .065 .174 .000 .089 .000 .079 .082 5.798 8.948 5.781 5.341 7.804 2.009 8.935 8.892 2.043 1.364 2.624 7.020 0.831 3.183 1.429 .997 .989 .959 .968 .980 .995 .990 1.000 .967 .992 .993 1.000 .993 .984 .967 .984 .993 1.000 .996 1.000 .995 .984 .962 .965 .997 1.000 .980 .988 .991 1.000 .897 .834 .813 .881 5 .023 5 .046 5 .034 5 .017 5 .030 5 .012 5 .026 5 .038 5 .015 2 .013 2 .024 2 .047 2 .010 2 .037 .018 2 .065 516.020 220 .069 455.372 220 p- value .804 CLF(Total) .326 .744 CLF(USA) .111 .796 CLF(China) .328 .903 LikC(Total) .376 .901 LikC(USA) .167 .904 LikC(China) .848 .861 LikW(Total) .112 .851 LikW(USA) .113 .870 LikW(China) .843 .837 LikS(Total2) .506 .846 LikS(USA2) .269 .838 LikS(China2) .030 .853 Discomfort(Total2) .660 .800 Discomfort(USA2) .204 .866 Discomfort(China2) .489 Parallelism (Total) NA .000 Parallelism (Total2) NA .000 Note. (CLF = collective face concerns; LikC = liking towards the critic; LikW = liking towards the noncritical witness; LikS = positive affect - liking towards the study; Discomfort = negative affect - feelings of discomfort for the study) These labels apply to Table 2 as well. CFAs for liking towards the study and discomfort feelings were conducted only among the four experimental conditions. Since the negative messages in the control conditions are not specific towards the participants’ countries, participants were not supposed to experience either positive or negative affect. Hence, CFAs for these two measures were conducted just for the four experimental conditions. Total = All six conditions in total, N = 215; USA = American conditions in total, n = 100; China = Chinese conditions in total, n = 115; Total2 = Four experimental conditions in total, n = 160; USA2 = American experimental conditions in total, n = 76; China2 = Chinese experimental conditions in total, n = 84. Parallelism between all measures were also tested by CFA via LessR. As for the total sample including all six conditions, the average absolute residual w/o the diagonal is 0.051. As for the four experimental conditions in total, the average absolute residual w/o the diagonal is 0.054. 46 Table 2: Descriptive Statistics for Conditions CLF LikC LikW LikS Discomfort 1 2 3 4 5 6 7 8 9 M SD M SD M SD M SD M SD 3.949 1.164 4.097 1.449 4.703 0.925 5.193 1.000 2.571 1.227 3.727 1.199 4.283 1.183 4.756 0.846 5.165 1.047 2.482 1.210 4.067 1.053 4.400 1.052 4.475 1.100 NA NA NA NA 5.006 0.998 3.718 1.368 4.594 0.869 5.191 0.821 3.110 1.378 4.940 1.296 3.396 1.441 4.384 0.984 4.975 1.225 3.560 1.393 4.755 1.195 4.374 1.220 4.252 1.233 NA NA NA NA 3.886 1.150 4.246 1.248 4.670 0.936 4.910 1.181 3.755 1.410 4.410 1.026 NA NA NA NA NA NA NA NA 3.829 1.180 4.197 1.307 4.732 0.877 5.178 1.019 2.523 1.210 10 4.967 1.178 3.526 1.412 4.469 0.939 5.063 1.080 3.378 1.396 In 4.470 1.201 3.910 1.412 4.649 0.893 5.192 0.910 2.837 1.322 Ou 4.393 1.386 3.796 1.396 4.552 0.938 5.060 1.146 3.074 1.413 Co 4.455 1.177 4.386 1.139 4.349 1.171 NA NA NA NA T 4.434 1.272 3.983 1.356 4.531 0.991 NA NA NA NA T2 4.426 1.306 3.845 1.400 4.594 0.917 5.117 1.049 2.972 1.376 Note. Table 2 showed unadjusted raw means for measured variables. Since the negative messages in the control conditions are not specific towards the participants’ countries, participants were not supposed to experience either positive or negative affect. Therefore, as for liking towards the study and discomfort feelings, the data were only reported for the experimental conditions. 1 = American ingroup condition, n = 35; 2 = American outgroup condition, n = 41; 3 = American control condition, n = 24; 4 = Chinese ingroup condition, n = 34; 5 = Chinese outgroup condition, n = 50; 6 = Chinese control condition, n = 31; 7 = American conditions in total, n = 100; 8 = Chinese conditions in total, n = 115; 9 = American experimental conditions in total, n = 76; 10 = Chinese experimental conditions in total, n = 84; In = Ingroup conditions in total, n = 69; Ou = Outgroup conditions in total, n = 91; Co = Control conditions in total, n = 55; T = All six conditions in total, N = 215; T2 = Four experimental conditions in total, n = 160 47 Table 3: Ordinary Least Squares Regression Results to Predict Feelings of Discomfort F (7, 152) = 5.929, p < .001, R2 = .214, adj. R2 =.178 Predictors DC Condition 1 DC Condition 2 DC Condition 3 C_CLF 1 x C_CLF 2 x C_CLF 3 x C_CLF Note. n = 160. B stands for unstandardized coefficients; β stands for standardized coefficients. t-scores 0.477 0.542 2.648 0.080 1.389 2.063 1.590 B 0.153 0.183 0.786 0.015 0.343 0.588 0.365 β .049 .055 .266 .014 .168 .219 .201 p-values .634 .589 .009 .936 .167 .041 .114 Reference group = The American ingroup-critic condition; DC Condition 1 = Dummy-coded American outgroup-critic condition; DC Condition 2 = Dummy-coded Chinese ingroup-critic condition; DC Condition 3 = Dummy-coded Chinese outgroup-critic condition; C_CLF = Centered Collective Face; 1 x C_CLF = The interaction term between centered collective face and dummy-coded American outgroup-critic condition; 2 x C_CLF = The interaction term between centered collective face and dummy-coded Chinese ingroup-critic condition; 3 x C_CLF = The interaction term between centered collective face and dummy-coded Chinese outgroup-critic condition. Centered collective face was calculated excluding the two control conditions. 48 REFERENCES 49 REFERENCES Back, K. W. (1951). Influence through social communication. The Journal of Abnormal and Social Psychology, 46, 9-23. doi:10.1073/h0058629 Bown, P. C. (2019, published online). The 2018 US-China trade conflict after forty years of special protection. China Economic Journal. doi:10.1080/17538963.2019.1608047 Brown, P., & Levinson, S. C. (1987). Politeness: Some universals in language usages. Cambridge, UK: Cambridge University. Cai, D. A., & Wilson, S. R. (2000). Identity implications of influence goals: A cross-cultural comparison of interaction goals and facework. Communication Studies, 51, 307-328. doi:10.1080/10510970009388529 Cialdini, R. B., Borden, R. J., Thorne, A., Walker, M. R., Freeman, S., & Sloan, L. R. (1976). Basking in reflected glory: Three (football) field studies. Journal of Personality and Social Psychology, 34, 366-375. doi:10.1037/0022-3514.34.3.366 Crocker, J., Thompson, L. L., McGraw, K. M., & Ingerman, C. (1987). Downward comparison, prejudice, and evaluations of others: Effects of self-esteem and threat. Journal of Personality and Social Psychology, 52, 907-916. doi:10.1037/0022-3514.52.5.907 Elliot, A. J., & Devine, P. G. (1994). On the motivational nature of cognitive dissonance: Dissonance as psychological discomfort. Journal of Personality and Social Psychology, 67, 382-394. doi:10.1037/0022-3514.67.3.382 Fein, S., & Spencer, S. J. (1997). Prejudice as self-image maintenance: Affirming the self through derogating others. Journal of Personality and Social Psychology, 73, 31-44. doi:10.1037//0022-3514.73.1.31 Festinger, L., & Carlsmith, J. M. (1959). Cognitive consequences of forced compliance. The Journal of Abnormal and Social Psychology, 58, 203-210. doi:10.1037/h0041593 Gleaves, J (2015). Manufactured dope: How the 1984 US Olympic cycling team rewrote the rules on drugs in sports, The International Journal of the History of Sport, 32, 89-107. doi:10.1080/09523367.2014.958667 Goffman, E. (1967). Interaction ritual: Essays on face-to-face behavior by Erving Goffman. New York, NY: Pantheon Books. Hail, H. C. (2015). Patriotism abroad: Overseas Chinese students’ encounters with Criticisms of China. Journal of Studies in International Education, 19, 311- 326. doi:10.1177/1028315314567175 50 Heng, T. T. (2019, published online). Understanding the heterogeneity of international students’ experiences: A case study of Chinese international students in U.S. universities. Journal of Studies in International Education. doi:10.1177/1028315319829880 Ho, D. Y. F. (1976). On the concept of face. American Journal of Sociology, 81, 867-884. doi:10.1086/226145 Ho, D. Y. F., Peng, S., Lai, A. C., & Chan, S. F. (2001). Indigenization and beyond: Methodological relationalism in the study of personality across cultural traditions. Journal of Personality, 69, 925-953. doi:10.1111/1467-6494.696170 Hogg, M. A., Terry, D. J., & White, K. M. (1995). A tale of two theories: A critical comparison of identity theory with social identity theory. Social Psychology Quarterly, 58, 255-269. doi:10.2307/2787127 Hu, H. C. (1944). The Chinese concepts of “face.” American Anthropologist, 46. 45-64. Retrieved from http://www.jstor.org/stable/662926 Hwang, K-K. (2012). Foundations of Chinese psychology: Confucian social relations (Electronic version). doi:10.1007/978-1-4614-1439-1 Koehn, H. P. (2008). Underneath Kyoto: Emerging subnational government initiatives and incipient issue-building opportunities in China and the United States. Global Environmental Politics, 8 (1), 53-77. doi:10.1162/glep.2008.8.1.53 Lyer, A., Schmader, T., & Lickel, B. (2007). Why individuals protest the perceived transgressions o their country. Personality and Social Psychology Bulletin, 33, 572-587. doi:10.1177/0146167206297402 Mackie, D. M., Devos, T., & Smith, E. R. (2000). Intergroup emotions: Explaining offensive action tendencies in an intergroup context. Journal of Personality and Social Psychology, 79, 602-616. doi:10.1037/0022-3514.79.4.602 Oetzel, J, Ting-Toomey, S, Masumoto, T, Yokochi, Y, Pan, X, & Takai, J. (2001). Face and facework in conflict: A cross-cultural comparison of China, Germany, Japan, and the United States. Communication Monographs, 68, 235–258. doi:10.1080/03637 750128061 Ohbuchi, K., Fukushima, O., & Tedeschi, J. T. (1999). Cultural values in conflict management: Goal orientation, goal attainment, and tactical decision. Journal of Cross-Cultural Psychology, 30, 51–71. doi:10.1177/0022022199030001003 Park, H. S., & Guan, X. (2006). The effects of national culture and face concerns on intention to apologize: A comparison of the USA and China. Journal of Intercultural Communication Research, 35, 183-204. doi:10.1080/17475750601026933 Shepherd, L., Spears, R., & Manstead, A. S. R. (2013). ‘This will bring shame on our nation’: The role of anticipated group-based emotions on collective action. The Journal of Experimental Social Psychology, 49, 42-57. doi:10.1016/j.jesp.2012.07.011 51 Smith, E. R., & Henry, S. (1996). An in-group becomes part of the self: Response time evidence. Personality & Social Psychology Bulletin, 22, 635-642. doi:10.1177/0146167296226008 Smith, E. R., & Mackie, D. M. (2015). Dynamics of group-based emotions: Insights from intergroup emotions theory. Emotion Review, 7, 349–354. doi:10.1177/1754073915590614 Svoboda, S. (2015). Why do so many Chinese students choose US universities? British Broadcasting Corporation. Retrieved from http://www.bbc.com Ting-Toomey, S., Gao, G., Trubisky, P., Yang, Z., Kim, H. S., Lin, S.-L., & Nishida, T. (1991). Culture, face maintenance, and styles of handling interpersonal conflict: A study in five cultures. International Journal of Conflict Management, 2, 275–296. doi:10.1108/eb022702 van Harreveld, F., Rutjens, B. T., Rotteveel, M., Nordgren, L. F., & van der Pligt. (2009). Ambivalence and decisional conflict as a cause of psychological discomfort: Feeling tense before jumping off the fence. Journal of Experimental Social Psychology, 45, 167- 173. doi:10.1016/j.jesp.2008.08.015 Yang, D. L., & Leung, A. (2008). The politics of sports anti-doping in China: crisis, governance and international compliance. China: An International Journal, 6, 121-148. Retrieved from http://link.galegroup.com.proxy2.cl.msu.edu.proxy1.cl.msu.edu/ apps/doc/A176775932/AONE?u=msu_main&sid=AONE&xid=a6e9839d Zhai, X. (2011). Perspectives on Chinese “Face”: Psychological motives and social representations. Beijing: Peking University (in Chinese). Zhu, Y. (2014, May). Face, its correlates, “reticence”, and alienation: The Chinese international students in the U.S. classroom. Paper presented at the annual conference of International Communication Association, Seattle, WA. Zhu, Y., & Bresnahan, M. (2018). Collective face, politeness strategies, and discomfort: Communication of American domestic students and Chinese international students. Journal of Intercultural Communication Research, 47, 141-159. doi:10.1080/ 17475759.2018.1438304 52