Geysers or Bubbling Hot Springs? A Cross-Cultural Examination of Customer Rage From Eastern and Western Perspectives

Abstract

There is compelling evidence that incidents of customer rage are on the rise and not just in Western, individualistic societies. Changing social and economic conditions in collectivist societies (e.g., emerging consumerism and rising wealth in China) have spawned the emergence of customer rage in Eastern countries. To this end, we examine how customer rage–associated emotions, expressions, and outcomes differ across Eastern and Western cultures. Results drawn from 982 frontline service customers from two Eastern (China and Thailand) and two Western (Australia and United States) countries show that although consumers from Western cultures are more likely to exhibit rage emotions, consumers from Eastern cultures are more likely to report a desire for revenge and express their emotions in both overt and more subtle ways. That is, customers from Eastern cultures, while slow to display anger, once initiated, their rage expressions toward frontline service workers tend to be physical and vengeful. This result contradicts generally held beliefs that Eastern consumers are reticent to express negative emotions. With these findings in mind, and to minimize damaging customer rage incidents, firms regularly engaged in cross-cultural service encounters need to provide intercultural awareness and communications training that equip frontline employees to understand customs, manners, etiquette, and expectations of Eastern and Western cultures in general and even specific cultural segments.

Keywords

anger rage culture cross-cultural service failure negative affect

Introduction

Extreme forms of customer aggression such as customer rage episodes are now an escalating phenomenon that affect up to two thirds of service customers and cost U.S. firms over US$200 billion in repeat sales (Customer Rage Study, 2015). Further, customer rage is an international phenomenon occurring across the globe in stores, online, and over the phone. Rage episodes have serious consequences not only for frontline employees facing the brunt of the outburst but also for the brand and even other customers. Further cause for concern arises from revenge behaviors against the offending firm perpetrated by enraged customers, which are responsible for substantial economic and social costs (Freeman 2013; Grainer et al. 2014; Grégoire, Tripp, and Legoux 2009). Indeed, some Fortune 500 firms are so concerned about the customer rage phenomenon they are installing software in call centers to detect rising levels of anger in customers’ voices, so a supervisor may intervene to prevent an incident escalating to rage. In today’s global marketplace, rage incidents are not confined to Western, individualist societies like the United States, Canada, France, or Australia. Nor should we assume that managerial practice naturally transfers across cultures. Baker, Meyer, and Chebat (2013), Chan, Yim, and Lam (2010), Gertson and Soderberg (2011), Patterson, Cowley, and Prasongsukarn (2006), Zourrig, Chebat, and Toffoli (2009), and others have shown that culture has a pervasive impact on, for example, brand positioning, interfirm collaboration, perceptions of service quality and value, revenge-seeking behavior, service recovery tactics, and customers’ preferences for personalized service.

While customers’ visible complaining behavior and emotional venting like anger are less common in Eastern, collectivist societies (Triandis 1995) such as China, Thailand, Vietnam, and Indonesia, there is evidence that rage incidents are on the rise in those countries (Patterson et al. 2009; Mummalaneni 2007; Surachartkumtonkun, Patterson, and McColl-Kennedy 2013). Recent world events support this view. For example, in Jakarta, Indonesia, passengers, after repeated frustrations with employees of a budget airline, smashed the front office at Jakarta’s Soekarno-Hatta airport (Febrina 2009). This trend has been attributed to a rapid rise in wealth (e.g., in China) and accompanying deep social changes and the alienation that many feel about the rate of change (Coonan 2006). Culture provides a normative framework that prescribes a range of socially acceptable behaviors. A key function of culture is to calibrate emotional responses, expressions, and behavior to culturally relevant situations to maintain social order. However, there is scant research in Eastern countries examining the manner to which emotions such as anger (whether overt or suppressed) manifest itself in rage expressions and, further, whether the pattern of behaviors is similar to that of Western consumers.

To begin to fill this gap, the broad purpose of this investigation is to examine the impact of national culture on how rage emotions are associated with various expressions of rage, and how this translates into behavioral outcomes toward the firm (revenge, word-of-mouth [WOM]). For example, Eastern consumers, due to social norms and display rules, are known to display less emotion than their Western counterparts when faced with an unfortunate situation (e.g., a service failure situation; Triandis 1995; Matsumoto, Yoo, and Chung 2010). However, while little is known about Eastern consumers’ subsequent emotional expressions and behavior, the general stereotype is that they are less demanding, aggressive, and vengeful when confronted with, for example, poor service recovery. Based on the hydraulic theory of anger and culture theory, we challenge this stereotype and suggest that Eastern customers are indeed more aggressive and vengeful than previously thought. Data for the study were drawn from respondents who experienced a rage incident in two very diverse cultural groups across four countries—Western, individualistic culture (United States and Australia) and Eastern, collectivist culture (China and Thailand). More generally, we examine cultural stereotypes across Eastern and Western countries.

Our study contributes to the literature in two important ways. First, as service encounters are first and foremost social exchanges between people, it stands to reason that cultural values and norms impact the regulation and expressions of emotions and subsequent conflict behaviors (Friedman, Chi, and Liu 2006). It therefore would be shortsighted to assume that displays of emotions as well as their accompanying expressions and manifestations transfer directly across cultures. Thus, the major contribution of this work is to examine the moderating role of culture. Second, this is one of only a few studies to examine the linkages between emotions, forms of rage expressions, and organization-directed behaviors.

Conceptual Development

Although both anger and rage are negatively valenced, high arousal, affective states, they are nonetheless distinguishable by their relative valences. As one author puts it, to say that rage and anger are the same emotion “is like saying a tornado is a bad storm” (Potter-Efron 2007, p. 9). In other words, rage is distinguishable from anger by its increased potency, vigor, and intensity. Compared to those who are just angry, enraged individuals tend to lack rationality, are more likely to lose control (Tavris 1989), and are more willing to enact revenge on an offending firm (Bechwati and Morrin 2003).

Research in behavioral psychology and organizational behavior has demonstrated that when individuals experience extreme forms of anger, they exhibit a tendency to want to attack the target verbally and/or nonverbally (Fitness 2000). These attacks are often unconscious, rather than premeditated reactions (Lord and Harvey 2002). Thus, extreme forms of anger can lead to overt behaviors that may result in damage to the organization’s property and/or emotional and physical harm to persons, including frontline employees, other customers, and even the customers themselves (Fullerton and Punj 1993; Gregoire and Fisher 2008). The overarching conceptual framework is shown in Figure 1.

Figure 1.

Conceptual framework.

The remainder of this article is organized as follows: (1) we discuss in more detail the nature of customer rage–associated emotions, expressions, and behaviors and how culture may moderate their development and execution; (2) develop a set of general hypotheses that guide the research; (3) outline the study method; (4) report the results of the data analyses and the extent to which the data support our hypotheses, before finally; (5) discussing the implications and directions for future research.

Extreme Anger, Expressions, and Organization-Directed Coping Behaviors

It is not surprising that anger is associated with the dissatisfaction emanating from a service failure, as the core relational theme of anger is other blame. Anger is a frequently experienced and powerful emotion that is known to be correlated with aggression, hostility, and retaliatory behavior (Bougie, Pieters, and Zeelenberg 2003). Moreover, recent research has shown anger, associated with a service failure, merges into rage over time if the issue is left unresolved (Surachartkumtonkun, McColl-Kennedy, and Patterson 2014). In this study, we employ a categorization of rage emotions and expressions that are based on the work of McColl-Kennedy et al. (2009), Shaver et al. (1987), and Deffenbacher et al. (2002). Specifically, we employ two dimensions of rage emotions (rancorous, retaliatory rage) and four dimensions of rage expressions: (1) verbal expression (e.g., yelling, swearing, name calling), (2) physical expression (e.g., making threats of physical harm, damaging property), (3) displaced aggression (taking anger out on others), and (4) nonverbal expression (rolling eyes, shaking head, dirty looks).

In keeping with established views on expressions of anger (e.g., Deffenbacher et al. 1996), the model is specified such that extreme forms of anger (retaliatory rage) are associated with extreme expressions (physical, verbal). In turn, the rage expressions are linked to specific customer retaliatory behaviors (see Figure 1). Whereas switching to a competitor and spreading negative word-of-mouth (NWOM) are rational behaviors for dissatisfied customers (DeWitt and Brady 2003; Keaveney 1995), enraged customers have the potential to engage in more serious revenge behaviors (Grégoire, Tripp, and Legoux 2009; Harris and Reynolds 2003). Revenge is distinguishable from retaliation and retribution, as the purpose of revenge is to give the avenger relief from anger. Moreover, it is seen as a last option to redress a sense of injustice (Zourrig, Chebat, and Toffoli 2009). Thus, in addition to WOM, we examine revenge as a second organization-directed behavior. Further, these emotions, expressions, and behaviors may be governed by societal norms and values that differ across cultural contexts. Next, we consider the effects of culture and justify its central role in our model.

Established Views on Culture and Anger Expressions

“Culture” is defined in this context as the sum of all behavioral norms and patterns collectively shared by a social group (Usunier 1996). Culture therefore provides essential boundaries that provide the acceptable range of emotional expression and behavior in social interactions (e.g., service encounters). Cultures calibrate the expressive and behavioral components of negative emotions such as anger via social roles and norms known as display rules. The calibration of emotions is essential for people in a given culture to exhibit behaviors consistent with social norms specified by culturally determined meanings in a given situation (Matsumoto, Yoo, and Chung 2010). Because service encounters are first and foremost social exchanges, it stands to reason that national culture, with its norms, expectations, and display rules, will impact customers’ expressions and behaviors when extreme anger is experienced following a dissatisfactory experience (Schoefer 2010).

Hofstede’s (1980, 1991) seminal work on the role of culture in shaping the norms of behavior in a society provide useful insights as to how consumers react to, and cope with, extreme negative emotions following an unsatisfactory service experience. Hofstede delineated four cultural dimensions: individualism/collectivism, uncertainty avoidance, power distance, and masculinity/femininity, with individualism/collectivism and power distance being relevant to the current study.

Individualistic cultures (e.g., United States, Australia, and Canada) promote the development of the independent self, autonomy, self-expression, and personal achievement (Markus and Kitayama 1991; Matsumoto, Yoo, and Chung 2010). The emphasis is on “I” rather than “We.” In contrast, collectivist cultures (e.g., China, Thailand, Vietnam, and Indonesia) foster interdependence and in-group goals, make a clear distinction between in-group (close friends, family) and out-group membership, seek group harmony, avoid public conflict, and place relatively more importance on societal norms as determinants of behavior. Public displays of emotion therefore are traditionally frowned upon. Individuals are expected to seek resolution to a conflict (e.g., an unsatisfactory service encounter) without losing their tempers. In Thailand, for example, the concept of “Jai Yen” or cool heart is a traditional value meaning people are slow to anger, and then propels Thais to mask any anger, even in the most stressful situations because to lose one’s temper is a sign of personal weakness (Triandis 1995; Patterson et al. 2009). In China and other Southeast Asian cultures, there is “mianzi,” or saving face, by maintaining surface tranquility (Johnson 2009).

Power distance is the second relevant cultural dimension. Power distance is the extent to which inequality between the more and less powerful members of society is considered acceptable and symbolizes status, respect, more rights, and wealth (Schutte and Ciarlante 1998). Cultures with a high power distance orientation (e.g., China, Thailand, and Indonesia) see “superiors” and “subordinates” on different social levels and consider differences in power and influence to be natural in society (Triandis 1995). Importantly, high power distance cultures make a clear distinction between in- and out-group membership. In turn, management rhetoric implies that the customer is sovereign and frontline employees are expected to do everything they can to satisfy customers, thus maintaining the status quo on power distance. Alternatively, low power distance cultures (e.g., United States, Australia, and Canada) believe people are equal and view inequalities in roles as established for the sake of convenience (Gudykunst, Ting-Toomey, and Chua 1988). Compared to high power distance cultures, customers in low power distance cultures will often downplay any perceived inequality when dealing with frontline employees (i.e., out-group).

Research has yet to examine how national culture might impact customer expressions and behaviors following a rage incident. Considerable evidence in social psychology (e.g., Berry et al. 1992; Venaik and Brewer 2010) indicates that psychological processes are culturally contingent. Culture is known to calibrate the expressive and behavioral components of negative emotions via social roles and behavioral norms referred to as display rules (Baker, Meyer, and Chebat 2013; Matsumoto, Yoo, and Chung 2010). Consequently, our understanding of how customers from different cultures evaluate service consumption and postconsumption experiences is limited. This research is intended as a launching point to begin to fill this gap.

Hypotheses Development

Rage-Associated Emotions

Following McColl-Kennedy et al. (2009), we examine two categories of rage-associated emotions: (1) rancorous rage (e.g., disgust, contempt) and (2) retaliatory rage (e.g., fury, wrath, ferocity), with the latter being more emotion laden. When customers experience rage, they either subconsciously elect to vent these feelings through physical, verbal, or nonverbal ways or choose to suppress their feelings. While coding schemas that link facial expressions to specific emotions have been developed by social psychologists (Ekman 1992), emotion regulation theory indicates that the intensity of emotions may be dampened or masked to conform to cultural norms (Matsumoto and Kupperbusch 2001) or the specific setting (in public vs. private) in which the emotion occurs. The practice of masking emotions has evolved in order to avoid offending the other party. Such a display rule helps foster harmony, cooperation, or cohesion in the relationship between parties.

Personal relationships have special saliency in Eastern, collectivist cultures. Schutte and Ciarlante (1998) and Triandis (1995) point to the value placed on conflict avoidance and the maintenance of smooth and harmonious interpersonal relationships. One of the main functions of culture is to moderate emotional responses to culturally relevant situations in order to maintain harmony. In collectivist societies, in the interests of maintaining interpersonal social harmony, anger is mostly neutralized or masked in accord with societal norms and display rules. Further, individuals often cope with negative events by believing “it was meant to be” (Cohen 1960). For example, the concept of mai bpen rai (“no worries,” “never mind”) is frequently invoked in Thailand when something unfortunate occurs, reflecting the notion that one should gracefully submit to forces beyond their control, in a similar manner to fate (McCarty et al. 1999). Thus, consumers from Eastern cultures are thought to be generally less prone to rage emotions than consumers from Western cultures. Moreover, customers from Western, individualistic cultures are considered to have higher expectations and be more sensitive to poor customer service (Mattila 1999). Accordingly, while not formally hypothesized, it is generally accepted that consumers from Western, individualistic cultures (United States and Australia) will feel more (a) rancorous rage and (b) retaliatory rage than their Eastern, collectivist counterparts in countries such as China and Thailand.

Rage Expressions and In-Group/Out-Group Cultural Differences

As previously noted, cross-cultural scholars generally hold the view that collectivist cultures tend to value emotional restraint and anger regulation more than individualistic cultures (Matsumoto, Yoo, and Chung 2010; Triandis 1995). Thus, overt expressions of anger are frowned upon in Eastern, collectivist, hierarchical cultures. Our prediction for how rage is expressed in response to extreme service dissatisfaction, however, is just the opposite. The logic for this prediction rests in (a) the way with which consumers in Eastern cultures are socialized to view self-in-group versus self-out-group members in the context of a service failure situation and (b) the hydraulic model of anger. Each is now discussed.

(a) In-groups and out-groups

To this point, our discussion of the cultural rules that govern expressions of rage has largely ignored context. Such an omission is perhaps a critical oversight in the context of frontline service encounters, as there are important differences between how in-group members are expected to treat out-group members across Eastern and Western cultures (Matsumoto, Yoo, and Chung 2010). In-groups are defined as groups of individuals about whose welfare a person is concerned, with whom that person is willing to co-operate without demanding equitable returns, and separation from whom leads to anxiety (Triandis 1995, p. 75). Moreover, in-groups are usually characterized by similarities among the members, where individuals share a sense of “common fate.” Family, close friends, and other close affinity groups (e.g., community-based religious groups) are typical of in-groups. Collectivists want their in-groups to be monolithic and homogenous—with everyone thinking, feeling, and acting in much the same way—and believe this leads to group harmony. Collectivists feel more comfortable than individualists in the presence of like-minded others.

Members of Eastern collectivist cultures tend to belong to relatively few social or in-groups but have considerably stronger in-group bonds compared to members of Western, individualistic cultures (Hui and Yee 1994; Triandis 1995). By contrast, in Western cultures, people tend to belong to more social groups but have relatively weak in-group bonds. However, their willingness to interact with strangers (out-group) is a feature of low power distance, egalitarian societies. Indeed, in such cultures, norms dictate civility when interacting with out-groups (Oyserman, Coon, and Kemmelmeier 2002).

The differences in social groups (in- vs. out-groups) have implications for how Eastern versus Western cultural members express anger to out-group members, such as frontline service workers (Allport 1953). Frontline employees would be considered out-group members by customers. Contrary to what might be expected, Eastern high power distance cultures are associated with relatively more anger toward out-groups, because doing so reinforces group identity and the social status quo (Matsumoto, Yoo, and Chung 2010). In the current study, unless a frontline employee was a close friend, member of a close affinity group, or family member of the aggrieved customer who went into rage, the employee could not be considered an in-group member. The chances of this are remote and indeed there was no evidence in the respondent-supplied narratives of the incident that this was so. It was assumed therefore that customers were treated as out-group members.

Thus, although rage expressions are thought to be less common in Eastern cultures, this general rule may not apply to expressions toward out-group members, such as frontline service workers. Conversely, whereas the expression of negative emotions is generally tolerated by members of low power distance, Western cultures, they are less likely to display negative emotions toward out-groups, “because doing so minimizes the differences between ingroups and outgroups and treats outgroup members like ingroup members in a relatively more egalitarian fashion” (Matsumoto, Yoo, and Chung 2010, p. 131).

(b) Hydraulic model of anger

A second reason to believe that forms of rage expressions may be prominent in Eastern cultures has to do with how rage develops and, specifically, the hydraulic model of anger (Bushman 2002). Rage behaviors in our study only emerged after the complainant had given the offending firm multiple opportunities to affect good service recovery. To illustrate, in one instance, a customer of an insurance firm made nine calls to a call center and made two personal visits to the firm’s headquarters over a 2-month period to receive a justified refund on a policy overpayment. The customer reported feelings of initial annoyance, then frustration, anger, and finally rage. Others explained how their negative feelings escalated when their calls to have the issue (service failure) resolved were delayed or even ignored. This is consistent with Ray, Wilhelm, and Gross’s (2008) finding that anger is maintained and even escalates when rumination occurs. With this in mind, we argue that customers eventually reach a boiling point after continued frustration in not achieving a just and fair resolution.

The hydraulic model of anger supports this; frustrations that lead to anger build up inside an individual in a similar way to hydraulic pressure inside a closed vessel, until it is released in some manner (Bushman 2002). Thus, if emotions are bottled up or masked in some way, such as is typically the case in Eastern cultures, they will eventually explode and lead to rage expressions. Or as expressed by Lakoff and Kovecses (1987, p. 201) who describe anger as heat in a container:

Anger is the heat of a fluid in a container … When the intensity of anger increases, the fluid rises … Intense anger produces pressure on the container … When anger becomes too intense, the person explodes, parts of him go up in the air …. When a person explodes, what was inside comes out.”

Recent world events support this view. Thailand is a country with 90% of the population claiming to adhere to Terevada Buddhist precepts that espouse harmony, conflict avoidance, and avoidance of open displays of negative emotions (Mulder 1996). Yet the street protests and demonstrations of May 2010 in Bangkok run counter to this view. After years of frustration due to perceived government inequity in aiding certain sections of society, many people reached boiling point, an “intolerable threshold,” resulting in physical conflict and loss of life (Economist 2010).

Moreover, there is emerging evidence that the justification for overt aggressive behavior (e.g., verbal, physical expressions) as a means of punishment has low acceptance in Western societies but relatively high moral approval in Asian countries (Ramirez 2007). And as noted earlier, the cultural norm of moral reciprocity is known to propel Asians to retaliate when wronged. Unlike spontaneous displays of emotion, reciprocity of certain behavior is considered acceptable and within social norms in Asian societies (Ramirez and Andreu 2003).

By contrast, it is generally accepted that the venting of negative emotions in public is widespread in the United States and Australia, although recent research suggests in some circumstances at least, exacting revenge may lead one to ruminate about the incident (Carlsmith, Wilson, and Gilbert 2008). In cartoons, movies, and plays, people are encouraged to “let off steam.” It is a form of catharsis and viewed as a way of relieving the pressure that feelings of annoyance and anger create inside the human psyche. The hydraulic model of anger would suggest that, under these conditions, rage expressions in Western cultures should be attenuated by the gradual letting off of steam.

In summary, the current investigation involves customers interacting with out-group members (frontline employees). In such circumstances, Eastern, high power distance, hierarchical cultures endorse, in the case of a service failure, relatively more hostility toward out-group members more so than in Western, low power distance cultures, in order to maintain the social hierarchy status quo. Moreover, as noted above, the fact that people in Eastern cultures tend to mask (or “bottle up”) their escalating negative emotions means that eventually they will exhibit more volatile verbal and nonverbal expressions of their anger than will their Western counterparts. Based on the preceding discussion:

Hypothesis 1: Rancorous rage emotions will be more strongly associated with verbal and nonverbal expressions in Eastern cultures than Western cultures.

Hypothesis 2: Retaliatory rage emotions will be more strongly associated with (a) physical expressions and (b) verbal expressions in Eastern cultures than in Western cultures.

Finally, as both Western and Eastern customers are known to also exhibit less aggressive forms of rage expressions (e.g., displaced anger), we make no predictions based on cultural differences.

Rage Expressions and Organization-Directed Behaviors

NWOM

Voicing NWOM to family, friends, and colleagues is viewed as a form of coping with a stressful event, and at the same time, a form of retaliation (Cropanzano 2001; Grégoire, Tripp, and Legoux 2009). It is typically motivated by a sense of injustice, especially in outward-focused situations (Barclay, Skarlicki, and Pugh 2005), and is known as a “cool and calculated response” (Cropanzano 2001, p. 201). NWOM, as a less aggressive form of organizational-directed behavior than revenge, should be positively associated with less aggressive forms of rage expressions (i.e., displaced and nonverbal expressions). From a cultural standpoint, Eastern collectivist cultures are characterized by the primacy of interpersonal communications within in-groups (Matsumoto, Yoo, and Chung 2010; Money, Gilly, and Graham 1998) and where the interests of the in-group take precedence over the desires and needs of the individual. A friendship prescribes the cooperative acts and relational norms of the group (Grayson 2007). As such, members of an in-group share information and place a high value on in-group communications networks and personal recommendations. Taken together with the primacy of in-groups for collectivist cultures, we expect an especially strong link to NWOM for Eastern consumers whose displaced and nonverbal expressions routes are activated. Hence, Hypothesis 3a and 3b:

Hypothesis 3: The associations between (a) displaced and (b) nonverbal expressions of rage and NWOM will be stronger in Eastern than in Western cultures.

Culture and Revenge

Typical of collectivist cultures is a common view—“don’t get angry, get even,” meaning that individuals do not tend to express their feelings publicly but rather seek revenge. Chinese in particular are known to perceive higher levels of unfairness in complaint situations because it inflicts money and time loss, and hence their perceptions of injustice are heightened (Zourrig, Chebat, and Toffoli 2009). Recent research shows that perceptions of severe injustice and organizational greed dangerously energize customers’ need for retaliation and revenge (Grégoire and Fisher 2008; Grégoire, Laufer, and Tripp 2010). In their review of Chinese history and culture, Lipman and Harrell (1990) show that the cultural characteristics of collective memory and moral reciprocity can propel people to an unrestrained quest for revenge when wronged in a personal or commercial context. Reciprocity (Bao) is a fundamental Chinese ethic. Reciprocity of actions (favor and wronged, reward and punishment) is considered almost a matter of fact like a cause-effect relationship. In contrast, revenge is typically frowned upon in Western cultures (Pepitone 1997).

Hypothesis 4: Consumers from Eastern cultures are more likely than their Western counterparts to engage in revenge behavior.

And as noted earlier, while customers from Eastern collectivist cultures may be slow to anger, when anger does reach boiling point (retaliatory rage), they can be expected to react with a vengeance.

Hypothesis 5: Extreme forms of anger (retaliatory rage) will be more strongly associated with revenge activities in Eastern cultures than in Western cultures.

The preceding hypotheses form the basis for our investigation at the intersection of culture and rage emotions, expressions, and behaviors. Our efforts to measure and test these effects and their interrelations are discussed next.

Method

Four countries were chosen to maximize west-east cultural differences, as indicated by Hofstede’s (1980) cultural indices. To illustrate, on the individualism-collectivism dimension, the United States and Australia have indices of 91 and 90, respectively (i.e., high individualism) versus China and Thailand of 18 and 20, respectively (i.e., high collectivism). On Hofstede’s power distance dimension, the United States and Australia rated 40 and 36, respectively, versus China and Thailand of 60 and 64, respectively. In other words, both the United States and Australia are highly individualist, low power distance cultures whereas China and Thailand represent highly collectivist, high power distance cultures. Next, we describe the steps taken to ensure pre- and postdata collection equivalence.

Cross-Cultural Considerations

Our cross-cultural empirical design followed Berry’s (1989) and Hult et al.’s (2008) schemas, which combine the emic and the etic approach to attitudinal and behavioral measurement. The four country samples were designed to be as similar as possible to minimize both errors of a sampling (Craig and Douglas 2000) and those of a semiotic (Arnould 1989) nature. With this approach, the authors first began the research in their common home culture (the United States and Australia) and then applied the constructs to the other cultures (imposed etic). The emic phase of the China and Thailand research took the form of 12 exploratory in-depth interviews in each country conducted by one of the authors with research assistants who were bilingual in Thai and English, or Chinese and English, respectively. The purpose of this phase was to establish construct equivalence (i.e., functional, conceptual, and category equivalence). The results were analyzed for commonality, which in this case was high. Finally, the study constructs were then adapted to achieve a higher level of equivalence, which would permit valid cross-cultural comparisons.

Next, linguistic equivalence of our measures was established via translation-back translation procedures. Finally, the instructions, sequence of questions, and interviewing techniques were standardized and pretested with 60 respondents (15 from each country) to ensure the English meaning of the instructions, constructs, phrases, and words were equivalent. Phase 2 was a quantitative phase consisting of survey interviews in the four countries.

Sampling and Data Collection

Professional marketing research firms were contracted to collect the data from the four countries. In an effort to achieve data collection equivalence, the same instructions and briefing were given to the data collection agencies in each country. The population of interest was customers who had experienced and had a good recollection of an extreme form of negative emotion accompanied by some form of rage behavior (e.g., yelling, slamming the telephone, threatening, cursing) following a service failure in the past 6 months. Respondents were asked to write a description of the events, actions taken by the offending organization, and their emotions accompanying the events that unfolded. This was done in order to help them recall the events more clearly and, in turn, answer the scaled questions concerning emotions, expressions, and retaliatory behavior more factually. Data collection was standardized by sampling only in major urban areas of the United States, Australia, China, and Thailand. The nonresponse rate was less than 11% in all countries, and hence nonresponse bias was not considered a problem. The final sample sizes used in our analyses were 228, 231, 303, and 220 for the United States, Australia, China, and Thailand, respectively.

Effort was also made to standardize the samples in each country. Table 1 shows the demographic profile of each country sample. Such stabilization of the sample profiles strengthens the internal validity of the study (Harpaz, Honig, and Coetsier 2002). To provide a more robust test of our hypotheses, we examined whether there were any significant differences between service failure types across our samples that might impact the results. Five service failure types that typically trigger rage incidents were identified from the literature (e.g., Keaveney 1995) and verified from our coding (e.g., core service failure, unresponsive employee behavior, inappropriate behavior, unethical behavior, and slow speed of service). The data were coded independently by three judges from respondents’ description of the service failure incident. Interjudge reliability was 87%, exceeding the accepted benchmark of 80% (Latham and Saari 1984). Results indicate consistency in failure types across samples. For example, the most prominent category, core service failure (failure to deliver the promised service, systems failure), represented around 50% of all incidents and there was no significant difference between east-west samples. Similar results were obtained for three of the four other failure types. There was, however, a small, but statistically significant difference for unresponsive employee behavior (lack of product knowledge, unresponsive to customer requests), which was reported by 23.3% of the Western sample and 19.2% of the Eastern sample (χ² < 0.05). Nonetheless, since all the cross-sample differences are either nonsignificant or relatively small, it seems reasonable to conclude that failure type has little, if any, impact on our results.

Table 1.

Demographic Profile of Each Country Sample.

	United States	Australia	China	Thailand
Mean age (years)	44	39	37	31
% Female	57	53	59	51
% With high school education	16.6	21	22	14.6
% With university education	69	58	53	74

Scales

The scales for the constructs comprising rage emotions, expressions of rage, and behaviors were sourced from the literature and were thus considered to have construct validity (refer to Appendix). To maintain source consistency, the scales for rage types, expressions, and behaviors (revenge and NWOM) were sourced from McColl-Kennedy et al. (2009). Construct reliability was supported for all scales in all four samples, with estimates ranging from .80 (retaliatory rage) to .89 (verbal expression) in the Australian sample, .80 (rancorous rage) to .95 (physical expression) in the United States sample, .74 (retaliatory rage) to .87 (rancorous rage, revenge) in the Thai sample, and .74 (rancorous rage) to .85 (NWOM) in the China sample. Construct reliabilities, standardized parameter estimates, and average variances extracted can be found in Table 2.

Table 2.

Results of the CFA.

Construct	Australia	United States	Thailand	China
Rancorous rage
Construct reliability	.82	.80	.87	.74
Average variance extracted	60%	49%	62%	43%
Parameter estimate range	.74–.80	.60–.75	.74–.84	.43–.80
Retaliatory rage
Construct reliability	.80	.85	.74	.79
Average variance extracted	57%	58%	42%	49%
Parameter estimate range	.67–.85	.72–.80	.41–.81	.65–.77
Verbal expression
Construct reliability	.89	.84	.85	.79
Average variance extracted	62%	52%	54%	43%
Parameter estimate range	.73–.87	.69–.77	.66–.80	.58–.74
Physical expression
Construct reliability	.88	.95	.84	.82
Average variance extracted	65%	79%	51%	54%
Parameter estimate range	.68–.90	.77–.97	.62–.78	.65–.84
Displaced expression
Construct reliability	.85	.82	.75	.78
Average variance extracted	65%	61%	50%	54%
Parameter estimate range	.71–.93	.70–.88	.65–.80	.61–.82
Nonverbal expression
Construct reliability	.84	.83	.84	.78
Average variance extracted	64%	56%	57%	48%
Parameter estimate range	.69–.86	.60–.83	.65–.85	.49–.80
Negative word-of-mouth
Construct reliability	.85	.82	.83	.85
Average variance extracted	73%	70%	71%	74%
Parameter estimate range	.83–.88	.73–.93	.82–.86	.81–.90
Revenge
Construct reliability	.87	.91	.87	.84
Average variance extracted	68%	72%	63%	58%
Parameter estimate range	.69–.92	.76–.91	.57–.87	.53–.92

Note. CFA model fit: χ²/df = 2.04, CFI = .92; RMSEA = .031–.035. CFI = comparative fit index; CFA = confirmatory factor analysis; RMSEA = root mean square error of approximation.

Common Method Variance

Both procedural precautions (Podsakoff et al. 2003) and empirical assessments (Malhotra, Kim, and Patil 2006) were used to address concerns about common method bias. From a procedural standpoint, the surveys were anonymous, a variety of scale end points were utilized, scale items were interspersed to avoid response sets, and the items underwent a series of pretests to assess clarity and ambiguity (Podsakoff et al. 2003). From a statistical standpoint, multiple tests were employed to examine whether common method variance biased the results. First, we conducted a Harmon one-factor test and a confirmatory factor analysis (CFA)-based one-factor test; both revealed a single factor did not adequately represent the items in the study. We then applied a marker variable assessment technique (Lindell and Whitney 2001). This involves the identification of the lowest (r = .003) and second lowest (r = .005) correlations among the manifest variables in the data set. Using these correlations as estimates of method bias in the data, we calculated a discounted correlation matrix using the more conservative method bias estimate (r = .005) and compared it to the unadjusted matrix. The marker variable correlation estimate is well below the .20 threshold (Malhotra, Kim, and Patil 2006). Hence, we conclude that method bias does not pose a significant risk to the interpretation of the data.

Scale Validation and Measurement Equivalence

The psychometric properties of all scales were tested simultaneously across all four samples via CFA. Results indicate the model fits the data very well (χ²/df = 2.04, comparative fit index [CFI] = .92, root mean square error of approximation [RMSEA] = .031–.035), particularly considering the complexity of the models and diversity of the samples. Next, all scales were subjected to a series of measurement invariance tests via multigroup CFA. The invariance testing process consisted of two independent steps. First, tests were performed to justify pooling the data from the Eastern (China and Thailand) and Western (the United States, Australia) samples; that is, measurement equivalence was assessed between the samples from China and Thailand and then a second test was performed to assess equivalence between the samples from Australia and the United States as a means to justify pooling the data into Eastern and Western samples, respectively. Second, a separate set of invariance tests was performed on the pooled Eastern and Western samples. This second set of invariance tests was conducted to justify making mean and parameter comparisons across the Eastern and Western samples.

Both test sequences included tests for configural, metric, scalar, and factor variance invariance as prescribed by Steenkamp and Baumgartner (1998). All eight constructs and 26 items were tested simultaneously throughout the two series of tests. Model fit across constrained and unconstrained models was evaluated with ▵χ², but particular emphasis was given to ▵CFI. This step was taken because, unlike Δχ², ΔCFI is robust to the limitations of goodness-of-fit χ² tests (Bryant and Satorra 2012). A value of ▵CFI < .01 is supportive of invariance across models (Cheung and Rensvold 2002).

Results of the within-sample configural invariance tests support configural invariance within the Eastern (i.e., China vs. Thailand) and the Western (i.e., the United States vs. Australia) samples. Model fit was good for both the Western (χ² = 994/542 df, CFI = .94, RMSEA = .039–.047) and Eastern (χ² = 1,215/542 df, CFI = .90, RMSEA = .045–.052) CFA models, all factor loadings were significant at p < .001, and in support of discriminant validity, there were significant changes in model fit when pairs of constructs were constrained to unity versus the fit of freely estimated models (Steenkamp and Baumgartner 1998). Full metric invariance was supported within the Western (▵χ² = 27.5/18 df, ▵CFI = .001) and Eastern (▵χ² = 54.5/18 df, ▵CFI = .006) samples. Likewise, full factor variance invariance was supported within both samples (Western: ▵χ² = 14/8 df, ▵CFI = .001; Eastern: ▵χ² = 20/8 df, ▵CFI = .002). Although full scalar invariance was not supported in either sample (Western: ▵χ² = 43/26 df, ▵CFI = .002; Eastern: ▵χ² = 433/26 df, ▵CFI = .062), partial scalar invariance was achieved in that at least 2 items per factor exhibited scalar invariance (all but six loadings had ▵CFI < .003). These findings are interpreted as support for pooling the data within the Western and Eastern samples.

The next series of equivalence tests was aimed at establishing invariance between the pooled Western (n = 459) and Eastern (n = 523) samples. Configural invariance was supported, as the model fit the data well (χ² = 1,388/542 df, CFI = .94, RMSEA = .037–.043), all factor loadings were significant at p < .001, and there were significant changes in model fit when pairs of constructs were constrained to unity versus the fit of freely estimated models. Full metric invariance (▵χ² = 44.1/18 df, ▵CFI = .002) and factor variance invariance (▵χ² = 17.7/8 df, ▵CFI = .000) were also supported across the Western and Eastern samples. Scalar invariance, which is required for mean comparisons across samples, was partially supported. Specifically, at least 2 items per factor exhibited scalar invariance (Steenkamp and Baumgartner 1998) and, when the noninvariant intercepts were freed, model fit was similar to that of the full metric invariance model (▵CFI = .008).

Results

The results displayed in Table 3 are instructive in comparing mean values for emotions, expressions, and behaviors across Eastern-Western cultures. The mean values were calibrated via latent means analysis (Byrne, Shavelson, and Muthén 1989). The Western sample was selected as the baseline group with latent means constrained to zero; means for the Eastern sample were freely estimated. Thus, results should be interpreted as whether the estimated latent factor means in the Eastern sample are significantly different from the baseline (Western) group. Raw means aggregated across Eastern and Western samples as well as within each country are provided for comparison purposes.

Table 3.

Raw Means and Latent Mean Analysis Results.

Dependent Factors	Raw Means (SE) Latent Mean East West Critical Ratios
Rancorous rage	2.63 (.047)	3.38 (.055)	–10.004***
Retaliatory rage	2.08 (.041)	2.34 (.054)	–3.584***
Verbal expressions	2.35 (.042)	2.42 (.052)	–1.080^ns
Physical expressions	1.49 (.030)	1.22 (.029)	6.466***
Displaced expressions	1.65 (.035)	1.55 (.039)	1.766^†
Nonverbal expressions	2.71 (.055)	2.57 (.059)	1.851^†
Revenge	1.62 (.038)	1.36 (.039)	4.772***
Negative word-of-mouth	3.35 (.058)	3.24 (.069)	1.194^ns

Note. Raw means are presented for reference purposes. Latent means were estimated for the Eastern sample; thus, all critical ratios are interpreted as deviations from the Western sample. ns = nonsignificant.

***p < .001. ^†p < .10.

While no formal hypotheses were advanced, the data support the view (Matilla 1999) that Western consumers exhibit more rancorous rage (M = 3.38) than their Eastern counterparts (M = 2.63; t = −10.00, p < .001). This results pattern was consistent across the Western (M_{Aus, USA} = 3.30, 3.45, respectively) versus Eastern (M_{Thai, China} = 2.52, 2.70) country samples. Similarly, Western consumers exhibited more retaliatory rage (M = 2.34) than their Eastern counterparts (M = 2.08; t = −3.58, p < .001) and, again, the pattern was consistent across the Western (M_{Aus, USA} = 2.31, 2.37) and Eastern (M_{Thai, China} = 1.97, 2.16) country samples. Further, the data show no difference in verbal expressions across cultural groups (p < .30). But Eastern consumers exhibited more physical rage expressions (M = 1.49) than Western consumers (M = 1.22; t = 6.47, p < .001) and this pattern was consistent across the Eastern (M_{Thai, China} = 1.56, 1.42) versus Western (M_{Aus, USA} = 1.19, 1.24) country samples.

Referring to our formal hypotheses, the results in Table 3 fully support Hypothesis 4. Specifically, Eastern consumers (M = 1.62) were more likely to engage in revenge activities than Western consumers (M = 1.36; t = 4.77, p < .001) and the results pattern was consistent across the Eastern (M_{Thai, China} = 1.77, 1.51) versus Western (M_{Aus, USA} = 1.33, 1.38) country samples.

Hypotheses 1, 2, 3, and 5 were tested using multiple groups structural equation modeling (refer Table 4). The structural model was tested at the item level and across east-west groups simultaneously. A control for the criticality of the service transaction (assessed by respondents) was added to the model across groups; the variable did not have a significant effect on any of the dependent variables and was therefore removed. The model fits the data reasonably well (χ²/df = 3.8, CFI = .88, RMSEA = .051–.056); specifically, while the absolute fit statistics (χ², RMSEA) are reasonable, incremental fit as reflected in CFI is only marginal.

Table 4.

Structural Model Results.

	Standardized Parameter Estimates		R ²
Model Relationships	Eastern Samples	Western Samples	East	West
Retaliatory rage → Physical expression^a	.529***	.389***	.280	.151
Rancorous rage → Verbal expression^a	.842***	.445***	.580	.377
Retaliatory rage → Verbal expression	–.08 ^ns	.206*
Rancorous rage → Displaced expression	.377***	.340***	.138	.116
Rancorous rage → Nonverbal expression^a	.733***	.497***	.526	.247
Retaliatory rage → Revenge^a	.326***	.229***	.379	.427
Verbal expression → Revenge	.218**	.07 ^ns
Physical expression → Revenge	.367***	.495***
Nonverbal expression → Revenge^a	–.237***	.011 ^ns
Rancorous rage → NWOM	.201 ^ns	.386***	.437	.462
Verbal expression → NWOM^a	.095 ^ns	.334***
Displaced expression → NWOM	–.071 ^ns	–.079 ^ns
Nonverbal expression → NWOM^a	.447***	.109*

Note. NWOM = negative word-of-mouth; ns = nonsignificant.

^aIndicates the parameter strength is significantly different across groups (p < .05).

*p < .05. **p < .01. ***p < .001.

Two caveats to model fit heuristics are worth considering here. First, as Cheung and Rensvold (2001) and others (e.g., Baumgartner and Steenkamp 1998) note, application of arbitrary cutoff values should be approached with caution when models are very complex, such as the model tested here, and when the samples are varied and diverse. In such cases, it is recommended that models be evaluated with lower cutoff values (see Cheung and Rensvold 2001, for a discussion of this issue). Second, the model tested is relatively parsimonious, which is a desired trait that unfortunately increases the likelihood of encountering parsimony error (Hall, Snell, and Foust 1999) and adversely impacts model fit estimates that do not account for parsimony, such as CFI. Taken together with the encouraging absolute fit statistics, we conclude that the structural model fits the data reasonably well. Parameter estimates were compared by constraining all paths to equality across samples, releasing the constraint on the path of interest, and comparing the change in model fit across the constrained and unconstrained models (Steenkamp, Batra, and Alden 2003). Model testing results, including standardized parameter estimates and R² values across east-west groups, are displayed in Table 4. A correlation matrix is shown in Table 5. Results are also provided across individual country samples for comparison purposes.

Table 5.

Correlation Matrices.

	Ranc	Retal	VerbEx	PhysEx	DispEx	NonvEx	Reveng	NWOM
Eastern sample
Ranc	—
Retal	.73	—
VerbEx	.50	.53	—
PhysEx	.35	.44	.69	—
DispEx	.17	.23	.49	.57	—
NonvEx	.52	.49	.83	.37	.33	—
Reveng	.42	.46	.45	.57	.26	.23	—
NWOM	.49	.35	.55	.19	.19	.65	.26	—
Western sample
Ranc	—
Retal	.74	—
VerbEx	.51	.53	—
PhysEx	.17	.39	.36	—
DispEx	.24	.36	.42	.58	—
NonvEx	.41	.35	.68	.29	.37	—
Reveng	.26	.47	.37	.61	.46	.27	—
NWOM	.60	.39	.60	.10	.21	.47	.27	—

Note. Ranc = rancorous rage; Retal = retaliatory rage; VerbEx = verbal expressions; PhysEx = physical expressions; DispEx = displaced expressions; NonvEx = nonverbal expressions; Reveng = revenge; NWOM = negative word-of-mouth.

Hypothesis 1a was partially supported in that, in general, the rancorous rage-verbal expressions link is stronger among Eastern consumers (γ = .842) than Western consumers (γ = .445). However, this effect proved to be very strong in the sample from Thailand (γ = .689) yet weak in the sample from China (γ = −.056). Thus, it is more accurate to say that the rancorous rage-verbal expressions relationship is significantly stronger in Thailand than in the other three countries. Hypothesis 1b was supported, as the effect of rancorous rage on nonverbal expressions was stronger in the Eastern sample (γ = .733) than in the Western sample (γ = .497). Again, this effect was especially strong in Thailand (γ = .925) versus the other three countries (γ = .388–.551).

The effect of retaliatory rage on physical expressions (Hypothesis 2a) is significantly stronger in the Eastern sample (γ = .529) than in the Western sample (γ = .389, p < .01), and this effect was consistent across the samples from Thailand/China (γ = .700 and .639, respectively) versus the samples from Australia/the United States (γ = .380 and .406). There was no significant difference between the effect of retaliatory rage on verbal expressions (Hypothesis 2b) across samples.

For Hypothesis 3, we argued that in Eastern cultures, the strength of the association between (a) displaced and (b) nonverbal expressions of rage and NWOM will be stronger than in Western countries. The data support Hypothesis 3b in that the link between nonverbal expressions and NWOM was stronger among Eastern consumers (β = .447 vs. .109). This effect holds across the samples from Thailand/China (β = .452 and .379, respectively) versus the samples from Australia/the United States (β = .208 and .164). Counter to our Hypothesis 3a, there were no statistically significant differences across cultural groups for the effect of displaced expressions on NWOM. A summary of hypotheses supported (and not supported) is shown in Table 6.

Table 6.

Summary of Results.

H1	Rancorous rage emotions will be more strongly associated with (a) verbal and (b) nonverbal expressions in Eastern cultures than Western cultures.	H1a partially supported H1b supported
H2	Retaliatory rage emotions will be more strongly associated with (a) physical expressions and (b) verbal expressions in Eastern cultures than Western cultures.	H2a supported H2b not supported
H3	The associations between (a) displaced and (b) nonverbal expressions of rage and NWOM will be stronger in Eastern than in Western cultures.	H3a not supported H3b supported
H4	Consumers from Eastern, collectivist cultures are more likely to engage in revenge than their Western counterparts.	H4 supported
H5	Extreme forms of anger (retaliatory rage) will be more strongly associated with revenge activities in Eastern cultures than in Western cultures.	H5 supported

Note. NWOM = negative word-of-mouth.

Finally, as predicted in Hypothesis 5, the retaliatory rage-revenge path is significantly stronger for Eastern (γ = .326) than Western consumers (γ = .229). This result was consistent across the samples from Thailand/China (γ = .351 and .332, respectively) versus the samples from Australia/the United States (γ = .239 and .262).

Mediation tests were performed in accordance with the methods prescribed by Iacobucci (2008) and Zhao, Lynch, and Chen (2010). Specifically, mediation equations were tested simultaneously via structural equation modeling (Iacobucci 2008) and the indirect effects were assessed with a bootstrapping procedure (Zhao, Lynch, and Chen 2010) conducted across 2,000 bootstrap samples. Key results include indirect only (full) mediation of the effect of rancorous rage on NWOM via nonverbal expressions in the Eastern sample and partial mediation in the Western sample. Moreover, the effect of retaliatory rage on revenge was partially mediated by physical expressions. These results support the general processes (depicted in Figure 1) by which rage emotions give way to expressions and then ultimately end in some form of organization-directed behavior.

General Discussion

The empirical results show that the linkages between rage emotions, expressions, and consequential organization-directed behaviors are calibrated by national culture. Several key findings warrant further discussion.

First, while not formally hypothesized, the results reaffirmed that consumers from Western cultures were more likely than their Eastern counterparts to experience rage emotions (retaliatory, rancorous) following a service failure. As sociocultural norms tend to govern the type of emotions felt, this is perhaps not surprising as Western consumers are known to have higher expectations and are less tolerant when they receive poor service (Mattila 1999). In Eastern cultures, the value placed on harmony, the belief in fate (things [negative events] were meant to be), and a tendency to attribute service failure to external, uncontrollable sources means Eastern customers mask their emotions. Ironically, in Eastern collectivist cultures, this tendency makes it problematic for frontline employees to detect escalating negative emotions and take proactive steps to diffuse the situation.

Second, in line with our prediction that might be considered counterintuitive (Hypothesis 2), the results clearly show that Eastern customers, though slow to anger, exhibit rage expressions toward frontline service workers that tend to be more physical (e.g., tried to harm the employee, made physical contact with an employee, threatened physical harm) and vengeful (caused physical damage to the organization’s property) than Western consumers. When emotions are pushed beyond a threshold, Eastern customers will eventually “explode” and express their anger more forcefully and aggressively than Westerners.

Third, the frontline service context here sheds further light on our findings. While Allport’s (1953) work recognized that attachment to one’s in-group does not necessarily require hostility to out-group members, this pattern of behavior alters when national culture, especially a high power distance culture, is superimposed on the relationships. One’s culture ascribes different meanings to self-in-group and out-group relationships (Triandis 1995). Eastern cultures encourage emotional displays toward in-group members that are designed to maintain harmony and mutual cooperation to a far greater degree than in Western cultures. However, the opposite applies for self-out-group relationships. In such cultures, the power distance cultural norm dictates that it is socially acceptable to display relatively more hostility toward out-group members who are lower in the social hierarchy, such as frontline workers, because doing so maintains the power distance status quo. In contrast, customers in low power distance cultures display less anger expressions toward out-groups because doing so minimizes the social distance between groups in society and is consistent with an egalitarian culture.

Fourth, consistent with Hypotheses 4 and 5, our results showed that consumers from Eastern cultures were, overall, more likely than Western consumers to engage in revenge activities following a rage incident and that the more aggressive forms of rage emotions (retaliatory) were a stronger driver of revenge among Eastern consumers. Revenge—an attempt to impose some form of suffering upon those who made one suffer—is a universal phenomenon. Moreover, prior research shows that perceptions of severe injustice and organizational greed dangerously energize customers’ need for revenge (Grégoire, Laufer, and Tripp 2010). However, people can sway under the impact of social norms. To illustrate, the notion of “getting even” pervades Eastern cultures, thereby encouraging revenge behavior. For example, Chinese are known to perceive higher levels of unfairness in complaint situations since it inflicts an economic loss, which may heighten their sense of injustice. In short, taking revenge when wronged has high moral approval in Eastern societies. On the other hand, revenge actions are typically frowned upon in Western cultures (Pepitone 1997).

Finally, as predicted (Hypothesis 3), nonverbal expressions were positively linked to NWOM, and much more strongly in the Eastern sample compared to the Western sample. An interesting, yet not hypothesized result was that verbal expression emerged as a strong driver of NWOM in our Western sample. As noted earlier, this suggests that not only do American and Australian complainants tend to “wear their heart on their sleeves,” but this verbal expression of anger is followed up by potentially damaging NWOM that has the potential to do significant brand damage as well as impact the psychological well-being of frontline employees who absorb the brunt of the rage. It may well be that verbal expressions (e.g., swearing, screaming), unlike displaced expressions, are insufficient to bring Western consumers back to a state of psychological equilibrium. Hence, NWOM takes place as a form of retaliation and coping behavior. While physical and verbal expressions are significant drivers of revenge for the Eastern sample, it appears that the less emotive nonverbal expressions result in less revenge seeking at least in Eastern cultures. As noted earlier, that when pushed beyond a certain threshold, Eastern consumers react more aggressively than Westerners. However, in line with their cultural values, below that anger threshold, they are unlikely to act aggressively.

Managerial Implications

Our results indicate that service recovery is a complex issue when working with consumers from a range of diverse cultures and hence our findings have important implications for organizations that have an international customer base or work across international boundaries.

First, a general theme that emerged is that while Eastern consumers may be slow to exhibit anger but, once angered, their rage expressions toward frontline service workers tend to be more overt, aggressive, and vengeful than Western customers. These findings run counter to commonly held stereotypes of consumers in Eastern cultures but are explained in terms of how consumers from Eastern, high power distance cultures are expected to treat members of out-groups and the fact that rage episodes tend to boil over if they are not gradually tempered. Such knowledge can assist service organizations: (1) recognize warning signals of the onset of customer rage, (2) develop intervention mechanisms designed to stall the progression of anger to rage, (3) implement prevention strategies that provide an outlet for anger and rage emotions, and (4) adapt such strategies for different cultural segments or international markets.

Second, our findings suggest specific strategies for managing customer rage in cross-cultural frontline service interactions. First, all frontline employees should receive intercultural awareness and communications training. This would equip them to understand general customs, manners, etiquette, and expectations of Eastern and Western cultures in general and even specific cultural segments if need be. People from Eastern collectivist cultures use a nonverbal communication style where most of the message is either in the situational context or internalized in the person; very little is in the coded, explicit message. To some extent, frontline workers may be expected to read the person’s mind. This, of course, means that detecting negative emotions and expressions are more subtle and so customer dissatisfaction is often difficult to assess. Hence, when dealing with Eastern cultures, employees need to be trained to be vigilant and observe contextual, nonverbal cues such as facial expressions (e.g., frowning, displays of disappointment) and body postures. Such expressions tend to be associated with escalating rage episodes. In Western cultures, communication tends to be more explicit and relies mostly on the spoken word, meaning that a buildup of rage emotions can be more easily detected.

Third, “face” is a fundamentally important concept in Eastern cultures, and understanding ways of coping with a loss of face and restoring face is central to avoiding rage episodes. In Eastern cultures, loss of face (being shamed) can have disastrous personal consequences and is avoided at all costs. In contrast, people from Western cultures pay relatively little attention to face (Irwin 1996). So extensive is attention to face in many Eastern countries that it is a dominant force in shaping effective interpersonal encounters and behavior. A sincere apology from a senior manager/supervisor is an important part of any communication behavior in hierarchical, power distance cultures where face saving is elevated to high levels of importance.

Fourth, when an incident has progressed to where expressions, such as rolling of eyes or raising of the voice, are evident, employees need to immediately recognize these as important signals of impending rage and an indication of more serious consequences and immediately involve a supervisor. While an aggrieved Eastern customer may have expressed strong negative emotions, employees should be trained to remain calm as cultural norms dictate the avoidance of strong displays of emotions (negative and positive) in public and conflict avoidance at all costs. Furthermore, employees should be trained to isolate the enraged customer from others in a shared service environment. Finally, managers must reconsider general cultural stereotypes concerning how consumers from Eastern and Western countries express anger. In contrast to generally held beliefs, consumers from Eastern consumers, once pushed beyond the brink, are more likely to exhibit aggressive rage expressions and seek revenge than their Western counterparts.

Limitations and Further Research

While our cross-sectional study across east-west country samples provides new insights into the manner in which culture moderates aggrieved customers emotions, expressions, and behaviors, it has certain limitations that should be addressed in future research. First, the cultural dimensions of collectivism/individualism and power distance were employed in our study to support various hypotheses. However, both dimensions are moderately correlated, that is, highly collectivist (Eastern) cultures are also typically high in power distance, thus creating possible confounding effects and raising issues related to internal validity. Future research might attempt to disentangle them and assess their separate effects using a video-based experiment in which power distance and collectivism are manipulated (or primed).

Second, our study does not explore how individual difference variables might add explanatory power to why some customers exhibit rancorous and others retaliatory rage. To help answer this question, other relevant individual differences such as desire for control, anger proneness (Spielberger 1999), attribution of blame (Mattila and Patterson 2004), and seeking redress (Chebat, Davidow, and Codjovi 2005) might be examined. It would also be insightful to study the relative impacts of rage emotions, expressions, and behaviors over time and on other important outcome variables, such as repurchase intentions, customer loyalty, and customers’ willingness to pay price premiums.

Third, while not examined in the current study, it is worth noting that in high power distance cultures, customers who perceive themselves to be of higher (or lower) status than the employee who is serving them may express their rage in a different manner. Furrer, Liu, and Sudharshan (2000) argue that in such cultures, it is important to distinguish between services relationships in which customers are seen as more powerful than service employees from those in which customers are viewed as less powerful. This might be explored in future research.

Fourth, although we advocate for caution in the indiscriminate application of model fit heuristics, the relatively low fit of the main structural model is a limitation that should be noted. Future research should consider ways to improve fit as well as model explanatory power. In particular, testing the models with more data or opting for a technique with lower sample requirements could be considered in future studies.

Finally, the current study implicitly equated country with culture. However, since both the United States and Australia have significant migration programs from Southeast Asian (collectivist) nations, it is fair to say that some respondents may well have come from Asian backgrounds and thus confounded the results to some extent. Future research might measure the cultural orientation of the individual respondent rather than assume country is synonymous with culture (Schoefer 2010).

Footnotes

Appendix

Table A1.

Customer Rage Scales.

	Customer Rage Emotions Scales
Rancorous rage
CONTEMPT	Feeling of despising; lack of respect or reverence for something
RESENTMENT	Feeling of indignant displeasure or persistent ill will at something regarded as a wrong insult or injury
SCORN	Open dislike and disrespect or derision often mixed with indignation; extreme disdain, contempt, or derision
Retaliatory rage
MALICE	Desire to see another experience pain, injury, or distress; deep-seated often unexplainable desire to see another suffer
FURY	Involves an overmastering destructive rage that can verge on madness; intense, disordered, and often destructive rage; feeling of extreme fierceness or violence
WRATH	Desire or intent to revenge or punish; strong vengeful anger or indignation; desire for punishment for an offense
	Customer Rage Expressions Scales
Verbal
SWOREALD	Swore/cursed aloud during my interaction with the service employee(s)
NEGALD	Made negative comments about the service employee(s) or organization aloud
INSULT	Made insulting remarks to the service employee(s)
YELLEMP	Yelled at the service employee(s)
RAISED	Raised my voice at the service employee(s)
Physical
PHYCONT	Made physical contact with a service employee in anger
PHYHARM	Tried to physically harm a service employee
PHYTHRT	Threatened to do physical harm to the service employee(s)
DMGCAUS	Tried to cause damage to the service organization’s property
Displaced
TOOKOTHR	Took my anger out on other people who were with me
YELLOTHR	Yelled at other people who were with me
TOOKLATE	Took my anger out on other people later on after the incident
Nonverbal
ROLLEYE	Rolled my eyes at the service employee(s)
DIRTYLK	Gave the service employee(s) dirty looks
GLARED	Glared at the service employee(s)
	Customer Rage Behaviors Scales
Revenge
RVGACT	Took actions to get revenge on the organization or its employee(s)
RVGCON	Considered ways to seek revenge against the organization or its employee(s)
SABOTHO	Thought about ways to sabotage the organization or its employee(s)
WOM
WOMORG	Made negative comments to others about the organization
WOMEMP	Made negative comments to others about the organization’s employee(s)

Note. WOM = word-of-mouth.

Acknowledgement

The authors are very grateful to Mary Jo Bitner, the AE and the three anonymous reviewers for their helpful comments through the preview process. We also gratefully acknowledge the Australian Research Council (ARC) for providing funding for this research (DP0664410).

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Allport

Gordon W.

(1953), “The Trend in Motivational Theory,” American Journal of Orthopsychiatry, 23 (1), 107–119.

Arnould

Eric J.

(1989), “Toward a Broadened Theory of Preference Formation and the Diffusion of Innovations: Cases from Zinder Province, Niger Republic,” Journal of Consumer Research, 16 (2), 239–267.

Baker

Thomas L.

Meyer

Tracy

Chebat

Jean-Charles

(2013), “Cultural Impacts on Felt and Expressed Emotions and Third Party Complaint Relationships,” Journal of Business Research, 66 (7), 816–822.

Barclay

Laurie J.

Skarlicki

Daniel P.

Douglas Pugh

(2005), “Exploring the Role of Emotions in Injustice Perceptions and Retaliation,” Journal of Applied Psychology, 90 (4), 629–643.

Baumgartner

Hans

Steenkamp

Jan-Benedict

(1998), “Multi-Group Latent Variable Models for Varying Numbers of Items and Factors with Cross-National and Longitudinal Applications,” Marketing Letters, 9 (1), 21–35.

Bechwati

Nada N.

Morrin

Maureen

(2003), “Outraged Consumers: Getting Even at the Expense of Getting a Good Deal,” Journal of Consumer Psychology, 13 (4), 440–453.

Berry

John W.

Poortinga

Ype H.

Segall

Marshall H.

Dasen

Pierre R.

(1992), Cross-cultural Psychology: Research and Applications. Cambridge, UK: Cambridge University Press.

Berry

John W.

(1989), “Imposed Etics-Emics-Derived Etics: The Operationalisation of a Compelling Idea,” International Journal of Psychology, 24 (6), 721–735.

Bougie

Roger

Pieters

Rick

Zeelenberg

Marcel

(2003), “Angry Customers Don’t Come Back, They Get Back: The Experience and Behavioral Implications of Anger and Dissatisfaction in Services,” Journal of the Academy of Marketing Science, 31 (4), 377–393.

10.

Bryant

Fred B.

Satorra

Albert

(2012), “Principles and Practice of Scale Difference Chi-Square Testing,” Structural Equation Modeling: A Multidisciplinary Journal, 19 (3), 372–398.

11.

Bushman

Brad J.

(2002), “Does Venting Anger Feed or Extinguish the Flame? Catharsis, Rumination, Distraction, Anger and Aggressive Responding,” Personality and Social Psychology Bulletin, 28 (6), 724–731.

12.

Byrne

Barbara M.

Shavelson

Richard J.

Muthen

Bengt

(1989), “Testing for the Equivalence of Factor Covariance and Mean Structures: The Issue of Partial Measurement Invariance,” Psychological Bulletin, 105 (3), 456–466.

13.

Carlsmith

Kevin M.

Wilson

Timothy D.

Gilbert

Daniel T.

(2008), “The Paradoxical Consequences of Revenge,” Journal of Personality and Social Psychology, 95 (6), 1316–1324.

14.

Chan

Kimmy W.

Yim

Chi K.

Lam

Simon

(2010), “Is Customer Participation in Value Creation a Double-Edged Sword? Evidence from Professional Financial Services across Cultures,” Journal of Marketing, 74 (3), 48–64.

15.

Chebat

Jean-Charles

Davidow

Moshe

Codjovi

Isabelle

(2005), “Silent Voices: Why Some Dissatisfied Consumers Fail to Complain,” Journal of Service Research, 7 (4), 328–342.

16.

Cheung

Gordon W.

Rensvold

Roger B.

(2002), “Evaluating Goodness-of-Fit Indexes for Testing Measurement Invariance,” Structural Equation Modeling: A Multidisciplinary Journal, 9 (2), 233–255.

17.

Cheung

Gordon W.

Rensvold

Roger B.

(2001), “The Effects of Model Parsimony and Sampling Error on the Fit of Structural Equation Models,” Organizational Research Methods, 4 (3), 236–264.

18.

Cohen

Jacob

(1960), “A Coefficient of Agreement for Nominal Scales,” Educational and Psychological Measurement, 20 (1), 37–46.

19.

Coonan

Clifford

(2006), “Anger Release Bars All the Rage in Stressed-Out China,” The Independent, August 8.

20.

Craig

Samuel

Douglas

Susan

(2000), International Marketing Research, 2nd ed. Chichester, England: Wiley.

21.

Cropanzano

Russell

(2001), Justice in the Workplace: From Theory to Practice. Mahwah, NJ: Lawrence Erlbaum.

22.

Customer Rage Study (2015), (accessed November 2015), [available at http://research.wpcarey.asu.edu/marketing/customer-rage-costs-202-billion-in-repeat-sales/].

23.

Deffenbacher

Jerry L.

Lynch

Rebekah S.

Oetting

Eugene R.

Swaim

Randall C.

(2002), “The Driving Anger Expression Inventory: A Measure of How People Express Their Anger on the Road,” Behaviour Research and Therapy, 40 (6), 717–737.

24.

Deffenbacher

Jerry L.

Oetting

Eugene R.

Lynch

Rebekah S.

Morris

Chad D.

(1996), “The Expression of Anger and its Consequences,” Behavioral Research Theory, 34 (7), 575–590.

25.

DeWitt

Tom

Brady

Michael K.

(2003), “Rethinking Service Recovery Strategies,” Journal of Service Research, 6 (2), 193–207.

26.

Economist (2010), Angels with Bloody Hands, (accessed April 16, 2010), [available at http://www.economist.com/node/15908365].

27.

Ekman

Paul

(1992), “An Argument for Basic Emotions,” Cognition and Emotion, 6 (3/4), 169–200.

28.

Febrina

Anissa S.

(2009), “Altitudes with Attitude,” The Jakarta Post, October 12, 25.

29.

Fitness

Julie

(2000), “Anger in the Workplace: An Emotion Script Approach to Anger Episodes between Workers and Their Superiors, Co-Workers and Subordinates,” Journal of Organizational Behavior, 21 (2), 147–162.

30.

Freeman

Debbie

(2013), “New Customer-Rage Study out for Holiday Shopping Season: We’ve Got More Problems & Anger, Despite Customer-Care Efforts,” (accessed November 2015), [available at http://wpcarey.asu.edu/news-releases/2013-11-26/new-customer-rage-study-out-holiday-shopping-season].

31.

Friedman

Ray

Chi

Shu-Cheng

Liu

Leigh A.

(2006), “An Expectancy Model of Chinese-American Differences in Conflict Avoiding,” Journal of International Business Studies, 37 (1), 76–91.

32.

Fullerton

Ronald A.

Punj

Girish

(1993), “Choosing to Misbehave: A Structural Model of Aberrant Consumer Behavior,” Advances in Consumer Research, 20 (1), 570–574.

33.

Furrer

Olivier

Liu

Ben S.

Sudharshan

(2000), “Service Quality Perceptions: Basis for Cross-Cultural Market Segmentation and Resource Allocation,” Journal of Service Research, 2 (4), 355–371.

34.

Gertson

Martine C.

Soderberg

Anne-Marie

(2011), “Intercultural Collaboration Stories: On Narrative and Analysis as Tools for Research in International Business,” Journal of International Business Studies, 42 (6), 787–804.

35.

Grainer

Marc

Noble

Charles H.

Bitner

Mary Jo

Broetzmann

Scott M.

(2014), “What Unhappy Customers Want,” Sloan Management Review, 51 (3), 31–35.

36.

Grayson

Kent

(2007), “Friendship versus Business in Marketing Relationships,” Journal of Marketing, 71 (4), 121–139.

37.

Grégoire

Yany

Laufer

Daniel

Tripp

Thomas M.

(2010), “A Comprehensive Model of Customer Direct and Indirect Revenge: Understanding the Effects of Perceived Greed and Customer Power,” Journal of the Academy of Marketing Science, 38 (6), 738–758.

38.

Grégoire

Yany

Tripp

Thomas M.

Legoux

Renaud

(2009), “When Customer Love Turns into Lasting Hate: The Effects of Relationship Strength and Time on Customer Revenge and Avoidance,” Journal of Marketing, 73 (6), 18–32.

39.

Grégoire

Yany

Fisher

Robert J.

(2008), “Customer Betrayal and Retaliation: When Your Best Customers Become Your Worst Enemies,” Journal of the Academy of Marketing Science, 36 (2), 247–261.

40.

Gudykunst

William B.

Ting-Toomey

Stella

Chua

Elizabeth

(1988), Culture and Interpersonal Communications. Newbury Park, CA: Sage.

41.

Hall

Rosalie J.

Snell

Andrea F.

Foust

Michelle S.

(1999), “Item Parcelling Strategies in SEM: Investigating the Subtle Effects of Unmodeled Secondary Constructs,” Organizational Research Methods, 2 (3), 233–256.

42.

Harpaz

Itzhak

Honig

Benson

Coetsier

Pol

(2002), “A Cross-Cultural Longitudinal Analysis of the Meaning of Work and the Socialization Process of Career Starters,” Journal of World Business, 37 (4), 230–244.

43.

Harris

Lloyd C.

Reynolds

Kate L.

(2003), “The Consequences of Dysfunctional Customer Behavior,” Journal of Service Research, 6 (2), 144–161.

44.

Hofstede

Gert H.

(1991), Cultures and Organizations: Software of the Mind. London, UK: McGraw Hill.

45.

Hofstede

Gert H.

(1980), Culture’s Consequences: International Differences in Work-Related Value. Newbury Park, CA: Sage.

46.

Hui

C. Harry

Yee

Candice

(1994), “The Shortened Individualism-Collectivism Scale: Its Relationship to Demographic and Work-Related Variables,” Journal of Research in Personality, 28 (4), 409–424.

47.

Hult

G. T. M.

Ketchen

David J.

Griffith

David A.

Finnegen

Carol A.

Gonzales-Padron

Tracy

Harmancioglu

Nukhe

Huang

Ying

Berk Talay

Tamer Cavusgil

(2008), “Data Equivalence in Cross-Cultural International Business Research: Assessment and Guidelines,” Journal of International Business Studies, 39 (6),1027–1044.

48.

Iacobucci

Dawn

(2008), Mediation Analysis. Thousand Oaks, CA: Sage.

49.

Irwin

Harry

(1996), Communicating with Asia: Understanding People and Customs. St Leonards, Australia: Allen and Unwin.

50.

Johnson

Michael

(2009), “Who Is the Angriest of All?” The Herald Tribune, June 13–14, 7.

51.

Keaveney

Susan M.

(1995), “Customer Switching Behavior in Service Industries: An Exploratory Study,” Journal of Marketing, 59 (2), 71–82.

52.

Lakoff

George

Kovecses

Zoltan

(1987), “The Cognitive Model of Anger Inherent in American English,” in Cultural Models in Language and Thought, Holland

Quinn

, eds. Cambridge, MA: Cambridge University Press, 195–221.

53.

Latham

Gary

Saari

Lise M.

(1984), “Do People Do What They Say? Further Studies on the Situational Interview,” Journal of Applied Psychology, 69 (4), 422–427.

54.

Lindell

Michael K.

Whitney

David J.

(2001), “Accounting for Common Method Variance in Cross-Sectional Research Designs,” Journal of Applied Psychology, 86 (1), 114–121.

55.

Lipman

Jonathan N.

Harrell

Steven

(1990), Violence in China: Essays in Culture and Counterculture. New York: State University of New York Press.

56.

Lord

Robert G.

Harvey

Jennifer L.

(2002), “An Information Processing Framework for Emotion,” in Emotions in the Workplace, Lord

R. G.

Klimoski

Kanfer

, eds. San Francisco, CA: Jossey-Bass, 115–146.

57.

Malhotra

Naresh K.

Kim

Sung S.

Patil

Ashutosh

(2006), “Common Method Variance in IS Research: A Comparison of Alternative Approaches and Reanalysis of Past Research,” Management Science, 52 (12), 1865–1883.

58.

Markus

Hazel R.

Kitayama

Shinobu

(1991), “Culture and the Self: Implications for Cognition, Emotion and Motivation,” Psychological Review, 98 (2), 224–253.

59.

Matsumoto

David

Yoo

Seung H.

Chung

Joanne

(2010), “The Expression of Anger Across Cultures,” in International Handbook of Anger, Potegal

Stemmler

Spielberger

, eds. New York: Springer Publishing, 195–137.

60.

Matsumoto

David

Kupperbush

Cenita

(2001), “Idiocentric and Aallocentric Differences in Emotional Expression, Experience, and the Coherence between Expression and Experience,” Asian Journal of Social Psychology, 4 (2), 113–131.

61.

Mattila

Anna

Patterson

Paul G.

(2004), “Service Recovery and Fairness Perceptions in Collectivist and Individualist Contexts,” Journal of Service Research, 6 (4), 336–346.

62.

Mattila

Anna

(1999), “The Role of Culture in the Service Evaluation Process,” Journal of Service Research, 1 (3), 250–261.

63.

McCarty

Carolyn A.

Weisz

John R.

Wanitromanee

Kanchana

Eastman

Karen L.

Suwanlert

Somsong

Chaiyasit

Wanchai

Band

Eve B.

(1999), “Culture, Coping and Context: Primary and Secondary Control among Thai and American Youth,” Journal of Child Psychiatry, 40 (5), 809–818.

64.

McColl-Kennedy

Janet

Patterson

Paul G.

Smith

Amy K.

Brady

Michael K.

(2009), “Customer Rage Episodes: Emotions, Expressions and Behaviors,” Journal of Retailing, 85 (2), 222–237.

65.

Money

R. Bruce

Gilly

Mary C.

Graham

John L.

(1998), “Explorations of National Culture and Word-of-Mouth Referral Behavior in the Purchase of Industrial Services in the United States and Japan,” Journal of Marketing, 62 (4), 76–87.

66.

Mulder

Niels

(1996), Inside Thai Society: Religion, Everyday Life, Change. Amsterdam, the Netherlands: The Pepin Press.

67.

Mummalaneni

(2007), “Antecedents and Consequences of “Ideal” Consumer Behavior—A Cross Cultural Comparison,” Journal of Consumer Marketing, 20 (4), 352–371.

68.

Oyserman

Daphna

Coon

Heather M.

Kemmelmeier

Markus

(2002), “Rethinking Individualism and Collectivism: Evaluation of Theoretical Assumptions and Meta-Analyses,” Psychological Bulletin, 128 (1), 3–72.

69.

Patterson

Paul G.

McColl-Kennedy

Janet R.

Smith

Amy K.

Zhi

(2009), “Customer Rage: Triggers, Tipping Points, and Take-Outs,” California Management Review, 52 (1), 6–28.

70.

Patterson

Paul G.

Cowley

Elizabeth

Prasongsukarn

Kriengsing

(2006), “Service Delivery Failure: The Moderating Impact of Individual Level Cultural Value Orientation on Perceptions of Justice,” International Journal of Research in Marketing, 23 (3), 263–277.

71.

Pepitone

Albert

(1997), “Nonmaterial Beliefs: Theory and Research in Cultural Social Psychology,” in The Message of Social Psychology: Perspectives on Mind in Society, McGarty

Haslam

S. A.

, eds. Oxford, UK: Blackwell, 252–267.

72.

Podsakoff

Philip M.

MacKenzie

Scott B.

Lee

Jeong-Yeon

Podsakoff

Nathan P.

(2003), “Common Method Bias in Behavioral Research: A Critical Review of the Literature and Recommended Remedies,” Journal of Applied Psychology, 88 (5), 879–903.

73.

Potter-Efron

Ronald

(2007), Rage: A Step-by-Step Guide to Overcoming Explosive Anger. Oakland, CA: New Harbinger Publications.

74.

Ramirez

J. Martin

(2007), “Justification of Aggression in Several Asian and European Countries with Different Religious and Cultural Background,” International Journal of Behavioral Development, 31 (1 Serial No. 51), 9–15.

75.

Ramirez

J. Martin

Andreu

Jose M.

(2003), “Aggression’s Typologies,” International Review of Social Psychology, 16 (3), 125–141.

76.

Ray

Rebecca D.

Wilhelm

Frank H.

Gross

James J.

(2008), “All in the Mind’s Eye. Anger Rumination and Reappraisal,” Journal of Personality and Social Psychology, 94 (1), 133–145.

77.

Schoefer

Klaus

(2010), “Cultural Moderation in the Formation of Recovery Satisfaction Judgements: A Cognitive-Affective Perspective,” Journal of Service Research, 13 (1), 52–66.

78.

Schutte

Hellmut

Ciarlante

Deanna

(1998), Consumer Behavior in Asia. Hampshire, England: Macmillan Business.

79.

Shaver

Philip

Schwartz

Judith

Kirson

Donald

O’Connor

Cary

(1987), “Emotion Knowledge: Further Exploration of a Prototype Approach,” Journal of Personality and Social Psychology, 52 (6), 1061–1086.

80.

Spielberger

Charles D.

(1999), State-Trait Anger Expression Inventory: Professional Manual. Odessa, FL: Psychological Assessment Resources.

81.

Steenkamp

Jan-Benedict

Batra

Rajeev

Alden

Dana L.

(2003), “How Perceived Brand Globalness Creates Brand Value,” Journal of International Business Studies, 34 (1), 53–65.

82.

Steenkamp

Jan-Benedict

Baumgartner

Hans

(1998), “Assessing Measurement Invariance in Cross-Cultural Research,” Journal of Consumer Research, 25 (1), 78–90.

83.

Surachartkumtonkun

Jirapon

McColl-Kennedy

Janet R.

Patterson

Paul G.

(2014), “Unpacking Customer Rage Elicitation a Dynamic Model,” Journal of Service Research, 18 (2): 177–192. doi:10.1177/1094670514556275.

84.

Surachartkumtonkun

Jirapon

Patterson

Paul G.

McColl-Kennedy

Janet R.

(2013), “Customer Rage Back-Story: Linking Needs-Based Cognitive Appraisal to Service Failure Type,” Journal of Retailing, 89 (1), 72–87.

85.

Tavris

Carol

(1989), Anger: The Misunderstood Emotion. New York: Simon and Shuster.

86.

Triandis

Harry C.

(1995), Individualism and Collectivism. Boulder, CO: Westview Press.

87.

Usunier

Jean-Claude

(1996), Marketing Across Cultures. London, England: Prentice Hall.

88.

Venaik

Sunil

Brewer

Paul

(2010), “Avoiding Uncertainty with Hofstede and GLOBE,” Journal of International Business Studies, 41 (8), 1294–1315.

89.

Zhao

Xinshu

Lynch

John G.

Chen

Qimei

(2010), “Reconsidering Baron and Kenny: Myths and Truths about Mediation Analysis,” Journal of Consumer Research, 37 (2), 197–206.

90.

Zourrig

Haithem

Chebat

Jean-Charles

Toffoli

Roy

(2009), “Consumers Revenge Behaviour: A Cross-Cultural Perspective,” Journal of Business Research, 62 (10), 995–1001.