Sources of Inducing Shame versus Anger at In-group Failure and Consumption Type

Abstract

Purpose: This research aimed at exploring the antecedents of feeling ashamed and anger when customers perceive the rightness of object of criticism induced from in-group failure triggered due to my mistake or others' mistake, and identifying the effects of shame and anger on customers' consumption type. Research design, data and methodology: This research used 2 (failure caused by my mistake versus failure caused by others' mistake) between- subjects design, and collected 353 data through on-line survey, and structural equation model of Amos 21.0 was used to verify the hypotheses developed by reviewing the past literature. Results: First, feeling anger motivates customers to choose compensatory consumption behaviors whereas shame leads people to choose adaptive consumption behaviors. Second, customer's feeling of shame and anger is depending on the perceived rightness of the criticism induced from the failure caused by my mistake or others' mistake. Conclusions: Marketers should notice that even shame and anger are included to negative emotions, customers who feel ashamed are different from customers who feel anger in view of approaching consumption. They should conduct their marketing focused on the adaptive consumption to ashamed consumers and do the marketing based on compensatory consumption to angry consumers.

1. Introduction

Emotion is one of the important factors affecting consumers' consumption motivation. Research has demonstrated that negative emotions would motivate individuals to increase the inclination to overeat or even buy some expensive brands which usually exceed their general level of consumption (Faber & Christenson, 1996). Although previous research has provided many behavioral perspectives on consumers' response to in-group failure, there have been little attention to emotion differences between my mistakes and other‘s mistakes triggering the failure. We focused on shame and anger which come from the in-group failure.

Kim and Gal (2014) compared adaptive consumption behaviors with compensatory consumption behaviors based on consumers' different assessments of self-deficit. That is to say, individuals would choose adaptive/compensatory consumption behaviors according to whether the deficits came from the difference between ideal and actual self or from the difference between ought and actual self (Choi, Wang, & Chen, 2018).

According to locus of control, meeting the in-group failure would make individuals feel different negative emotions, such as shame or anger (Rotter, 1966). The locus of control consists of internal and external locus of control (Rotter, 1966). Internal locus of control orientated individuals to have a better sense of control over their actions, losses and rewards while people who hold the external locus of control tend to attribute their actions, results to outside forces (Hoffman, Novak, & Schlosser, 2003). Which means people with different locus of control would feel different emotions and choose different consumptions. It is necessary to find out which kind of emotions could be aroused and whether these emotions could influence consumption types.

The purposes of this research are exploring how the criticisms induced from in-group failure influence consumers‘ negative emotions and how the emotions impact consumers‘ choice between adaptive and compensatory consumption.

2. Theoretical Background and Hypotheses Development

2.1. Two Types of Failure Based on the Locus of Control

2.1.1. Locus of Control

Individuals hold the tendency to explain their daily events depending on the locus of control (Rotter, 1966). Locus of control could be divided into internal and external locus of control. To be more specific, people with the external locus of control would perceive that their actions and outcomes are due to outside forces, not under their control (Pinto, Mansfield & Parente, 2004). It is typically perceived as the result of luck, chance, or fate; as they were under the control of powerful others, or the great complexity of the forces surrounding them. In contrast, individuals who hold an internal control consider that the mostly outcomes are consisted with their characteristics (Hoffman et al., 2003). In another world, compared to individuals who hold the external locus of control, the internal locus of control orientated people would attribute their outcomes or events to themselves. These people believe that they could control their outcomes in the daily life (Pinto, Mansfield, & Parente, 2004). That is to mean, individuals with different kinds of locus of control would adapt different approaches to defense the threat, such as failure.

2.1.2. Failure Caused by My and Others' Mistake

Figuring out the cause of the failure is the first and important thing for the most people. And the cause could be divided into my and others‘ mistake depending on the locus of control. If individuals experience the failure caused by their own mistake, they may attribute the failure to their characteristics, such as ability, efforts, rather than complain about their opponents or bad luck (Hoffman et al., 2003). That is to say, when you find the reason of failure is yourself, you would perceive the lace of self-ability (Pinto, Mansfield, & Parente, 2004). Then you would be disappointed in yourself (Tangney, 2002). What‘s more, you may show more intention to improve and enhance yourself by working hard or taking relevant online courses (Rahim & Patton, 2015).

However, if people attributes the failure to the other competitors, juries or even environments, then this situation could be regarded as failure taken place because of others‘ mistake. In the other word, others or even external context could make the result ended in the failure to you. What is more important, failure taken place because of others‘ mistake is easy to arouse angry emotion (Andrews, Brewin, Rose, & Kirk, 2000). And meeting the failure caused by others‘ mistake, people maybe just want to have a drink or talk with their friends rather than do something to enhance their business-related capabilities because people did not attribute the reason to their abilities (Choi, Wang, & Chen, 2018).

2.2. Shame and Anger

Research demonstrated that consumers‘ emotions could influence their decision-making (Agrawal, Han, & Duhachek, 2013). Consumers tend to increase positive feelings or decrease negative feelings by choosing different consumption behaviors (Keltner & Lerner, 2010). Then based on the attribution theory, different attributions could cause different emotional responses, such as shame or anger (Machleit & Mantel, 2001), which means consumers would choose different consumption behaviors according to feeling ashamed or anger.

2.2.1. Shame

For mostly, shame may be a kind of desire to escape or disappear (Tangney & Tracy, 2011). People would experience shame when they realize that they cannot achieve their ideal self (Baumeister, Stilwell & Heatherton, 1994). And the result of a self-evaluation, shame often appear along with a kind of ―being smaller‖, a sense of powerlessness or inadequacy (Tangney & Dearing, 2003). Individual who is caught in the ashamed state may adjust or behave in positive ways. What‘s more, shame, a kind of weakness of consumers‘ ideal self, could motivate them to improve themselves by adopting adaptive consumption (Rahim & Patton, 2015).

2.2.2. Anger

Anger is different from other negative emotions according to some cognitive dimensions (Roseman, Wiest, & Swartz, 1994). The key dimensions are certainty, control and responsibility. Anger occurs as a result of individuals‘ appraisals of high other-responsibility for negative events and high other-control over these negative events (Averill, 1983). That means, anger is a retrospective emotion, which tends to occur when people attribute a goal-incongruent event to external sources. Such an external attribution implies to blame someone else for an aversion situation (Weiner, 1985). Consumers would drink alcohol or buy something more than their consumption level to cope their anger emotions (Choi, Wang, & Chen, 2018).

2.3. Object and Two Kinds of Criticisms

Criticism is an assessment or correction of events in daily life and could arouse the recipients‘ negative emotions or motivate them to adopt defense actions (Kluger & DeNisi, 1996). Research has showed that criticism could influence vengeful, uncivil, aggressive and so on (Aquino, Tripp, & Bies, 2001). Criticisms could be divided into constructive criticism and destructive criticism (Baron, 1988). In the field of education, research has find that a kind of criticism which contains not only critical feedback but also the real recommendation to the problem could be called as constructive criticism (Bjorklund, Parente, & Sathianathan, 2004). While Baron (1988) named another kind of criticism as destructive criticism which just includes negative feedback. More specifically, constructive criticism pays more attention to analyze the problem and present the relevant solutions to solve the problems . However, destructive criticisms just focus on the negative comments about outcomes (Baron, 1988).

According to self-categorization theory (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987), person‘s selfcategorization is an identity classification, including individual identity and social identity. Individual identity refers to the subjective feeling of self when one does not consider himself/herself in society, and social identity involves a public image that a person expresses through social roles and relationships (Cheek & Briggs, 1982). In other word, individual identity refers to what kind of person define themselves as (―me‖). Social identity refers to the self-categorization of people relative to their group membership (―we‖), which refers to the self-concept of people as social group members (Leaper, 2011). People see themselves as members of a group, based on the unity of perception and action among group members (Stets & Burke, 2000).

Also, there are other members (―others‖) except me in the social group. So when we faced the criticism from others after a group failed, objects of others' criticism can be divided into my mistake and other member‘s mistake. When my group fails because of me, others' criticism will be regarded as criticizing me or my group. When my in-group fails because of my in-group members, the criticism from others will be regarded as the criticism of my in-group members and my group.

2.4. Rightness of Two Types of Criticism

Criticism refers to the negative feedback obtained from others (Baron, 1988). Individuals and social groups will inevitably be criticized in social competition. That is to say, individuals could classify these criticisms into constructive criticisms or destructive criticisms depending on how rightness the individuals perceived from these criticisms. Specially, as for, if I hold the primary cause of our in-group failure, observers may not only criticize me but also include my whole in-group. Observers‘ reactions to the failure can be different to the actor‘s reaction. The two reactions can be contrasted between observers and actors. Actors are biased because of their bias towards positive message about themselves. This bias is not shared by observers. Actors are biased, but observers are objective. They are more likely to form a poorer opinion by attributing the failure to an ulterior motive (Vonk, 2002). Then in these case, when the object of criticisms is me, I will agree with them and think these criticisms are right. Then these constructive criticisms may arouse stronger level of negative emotions of shame rather than anger to me because I also realized that I was the cause of our in-group failure and I could be better in the next time (Allred, 1999). However, when the object of criticisms is my whole in-group, I may disagree with them and think they were wrong. Then these destructive criticisms may produce anger rather than shame to me (Quigley & Tedeschi, 1996). So the hypothesizes are presented as follows:

H1-1: When observers' criticism to the in-group failure taken place because of my mistake is focused on myself, the perceived rightness of the criticism will positively affect feeling of shame.

H1-2: When observers' criticism to the in-group failure taken place because of my mistake is focused on myself, the perceived rightness of the criticism will negatively affect feeling of anger.

H2-1: When observers' criticism to the in-group failure taken place because of my mistake is focused on in-group, the perceived rightness of the criticism will negatively affect feeling of shame.

H2-2: When observers' criticism to the in-group failure taken place because of my mistake is focused on in-group, the perceived rightness of the criticism will positively affect feeling of anger.

When the failure caused by other in-group members, these in-group members and whole in-group would be criticized by observers. According to social identity theory, individuals hold a tendency to maintain their positive social identity, which means individuals may kick their in-group member who brings threat to their in-group out (Stets & Burke, 2000). That means when the object of criticisms is in-group member, individuals would disagree with them  because in their opinions, the members who caused the failure to their in-group has dismissed from their in-group. So these destructive criticisms would increase individuals' anger and decrease their feelings of shame. However, if the observers comment more on the whole in-group, individuals would agree with them because individuals think that their in-group did fail. So these constructive criticisms may arouse individuals' negative emotion of shame and decrease the feeling of anger. So, the hypothesizes are presented as follows:

H3-1: When observers' criticism to the in-group failure taken place because of other group member's mistake is focused on the other, the perceived rightness of the criticism will negatively affect feeling of shame.

H3-2: When observers' criticism to the in-group failure taken place because of other group member's mistake is focused on the other, the perceived rightness of the criticism will positively affect feeling of anger.

H4-1: When observers' criticism to the in-group failure taken place because of other group member's mistake is focused on in-group, the perceived rightness of the criticism will positively affect feeling of shame.

H4-2: When observers' criticism to the in-group failure taken place because of other group member's mistake is focused on in-group, the perceived rightness of the criticism will negatively affect feeling of anger.

2.4. Adaptive and Compensatory Consumption Behaviors

2.4.1. Adaptive Consumption Behaviors

Kim and Gal (2014) named the consumer behavior which related with direct problem solving as adaptive consumption behavior. That is to say, adaptive consumption behavior places emphasis on the functions and attributes of products or services. Because the motivation that consumers choose adaptive consumption is improving or enhancing themselves (Choi, Wang, & Chen, 2018). Then when people buy products/services to pursue long-term solutions to problems from the roots, rather than merely solving the current bad state, the consumer behavior is considered to be adaptive consumption (Kim & Gal, 2014).

2.4.2. Compensatory Consumption Behaviors

Compensatory consumption behaviors are just the one of the consumption behaviors related to the threat situations, especially when individuals experience an aversive state of discomfort from them and this kind of consumption behavior could provide symbolic values to consumers (Lee & Shrum, 2013). For example, if MBA students fail through a job interview compared to their peers, they could exhibit strong purchase intention to the products or brands that signaled success, such as luxury watches, to repair themselves (Wicklund & Gollwizer, 1981). As we mentioned before, people tend to engage in compensatory consumption as their defense mechanism when their aversive states are aroused to translate their attention from the negative moods (Mandel, Rucker, Levav, & Galinsky, 2017). That is to say, compensatory consumption is a way for consumers to solve their current bad emotion state. 

2.5. The Relationship between Emotions and Consumption Behaviors

Previous research has confirmed that emotions can affect how individuals process information and make decisions (Zillmann, 2015). That is, consumers are easier to engage in behaviors guided by negative emotions. Exactly, not only can negative emotions drive individuals to take the root cause of the problem into consideration and find solutions to deal with it, but also can motivate consumers to escape from the bad feelings as far as possible (Zillmann, 2015). Then after humans experiencing failures based on internal or external orientation, what takes effect on their consumption behaviors?

ㆍ Appraisal theorists further show that people use different coping strategies to reduce such negative emotions (Lazarus, 1991). Ashamed customers often think deeply about why he or she would be the reason for the failure, and they could also show more interesting in changing their disadvantage. Shame often comes out along with a feeling of ―being smaller‖, or of being powerlessness or inadequacy (Tangney & Dearing, 2003). Ashamed customers pay more attention to engage in adaptive consumption behaviors which means they focus on the quality and utility of products or services more than the products‘ or services' symbolic values.

ㆍHowever, anger as retrospective emotion helps people attribute a goal-incongruent event to external sources. Angry customers often use confrontative coping, that is, retaliatory behavior toward the blameworthy organization. Speak more clearly, when consumers attribute failure to others, they are easy to feel angry, so they may just want to change their negative emotions rather than think about how to improve themselves, because in their eyes, the questions come from others rather than their own ability or responsibility.

ㆍHence, in this research, formally

H5: Consumers who feel ashamed are more likely to choose adaptive consumption behaviors

H6: Consumers who feel angry are more likely to choose compensatory consumption behaviors

Figure 1: Research Model

3. Empirical Study

3.1. Preliminary Survey and Selection of Study Subject

In order to measure whether individuals would show different consumption behaviors between shame and anger experienced, this research should find the two kinds of products or services which could represent the two kinds of consumption behaviors: adaptive consumption behaviors and compensatory consumption behaviors. Depending on the previous studies, our research conducted preliminary test (Lisjak, Bonezzi, Kim, & Rucker, 2014; Rustagi & Shrum, 2018).

This research offered to 50 participants 10 kinds of services and products (protein powder, fitness center, Six Walnuts, basketball shoes, basketball, dumbbell, online course, basketball suit, book, pen). After getting through these ten kinds of services and products, participants need to answer two questions. They are, 1. ―if you failed the sports competition, which two kinds of services or products would satisfy your inner need which means you want to improve yourself or enhance your ability?‖, 2. ―if you failed the sports competition, which two kinds of services or products would fulfill your inner need which means you want to show your power or just want to change your negative emotions?‖ At the last, we used the frequency analysis, which indicated that fitness center (65.52%), dumbbell (55.17%) and basketball (34.48%) were more likely to be considered as adaptive products/services. And books (65.52%), pen (31.03%) and online course (25.86%) were more likely to be considered as compensatory consumption products/services.

Based on the results of the preliminary survey and the types of failure we designed, we selected fitness center and books as follow-up experimental products, because fitness center is the hottest product/service for college students who love sports. And, according to Rustagi and Shrum's (2018）research books (eg., National Geographic magazine) and pens are thought to be reminiscent of intelligence.

3.2 Measures of Configuration Concepts

3.2.1. Failure Situation Manipulation

Then to let individuals experience failure caused by my/others‘ mistake according to locus of control, this research developed two failure situations (Rotter, 1966).

First, the situation to prime failure caused by my mistake was developed as follows.

I am a senior in Tianjin University (TJU). I loved basketball and I am the captain of TJU basketball team. I am determined to make it to the final in this year's "ANTA Cup" National Basketball League. This is the last time I participate in a competition on behalf of TJU. Yesterday I was leading the TJU basketball team battle with Tianjin Academy of Fine Arts team with my jersey of No.1. However, I felt physically weak in second half and missed the ball pass by my teammates several times. We failed. I thought it was because of me that the entire team was disqualified from entering the finals. Finally, I heard audience said that “No. 1 should have a break in the second half. If without No. 1 TJU must have won”, “No.1 is not good at basketball” and other people even said “The students of TJU are poor in physical fitness”, “TJU is too weak”.

Second, the situation to prime failure cause by others‘ mistake was developed as follows.

I am a student of Tianjin University (TJU), I am proud to be a student of TJU. I loved basketball and I have been chosen to be a member of TJU basketball team. Yesterday our team was battle with Tianjin Academy of Fine Arts team at the "ANTA cup" National Basketball League. But we failed. In fact, I did well in the first half. However, my teammate A (No. 5) had a very high turnover error in the second half and he did not follow the tactical plan we discussed before the game. I gave him a look and let him pass ball to me but he ignored me. I feel not good. Finally, I heard audience said that “No.5 is not good at basketball”, “TJU would have won it without No.5”, “No.5 did not pass the ball to others. It is he who caused the team to lose.” But some audiences said, “TJU is too weak.”

Following each situation, after individuals reading these two kinds of situations, we explain the definition of locus of control by the sentences ―internal locus of control—ones‘ belief about the extent of ones‘ control of the situation; external locus of control—one‘s belief about the extent to which events in his/her life are authorized by environmental  factors." Then one question was used to check whether individuals were experienced in the two kinds of failure, "which locus of control in situation above do you think you are concentrating on or what is the cause of the failure?"

3.2.2. Measures of Rightness of Criticism (RCI/RCO/RCG)

Depending on the definition of constructive criticisms and destructive criticisms (Baron, 1988), this research developed each of 3 items to measure the rightness of criticisms which could be divided into object of I（RCI）, other in-group member (RCO) and the whole in-group (RCG). They are ―1. When I was criticized by observers, I thought the criticisms were acceptable.‖ ―2. When I was criticized by observers, I thought the criticisms were desirable.‖ ―3. When I was criticized by observers, I thought I should be criticized.‖ ―4. When other in-group members were criticized by observers, I thought the criticisms were acceptable.‖ ―5. When the other in-group members were criticized by observers, I thought the criticisms were desirable.‖ ―6. When the other in-group members were criticized by observers, I thought they should be criticized.‖ ―7. When my in-group was criticized by observers, I thought the criticisms were acceptable.‖ ―8. When my in-group was criticized by observers, I thought the criticisms were desirable.‖ ―9. When my in-group was criticized by observers, I thought my in-group should be criticized,‖ all of which were answered on 7-point scale ranging from [1]=not at all and [7]=very much.

3.2.3. Measures of Feeling of Shame(SH)

According to the definition of the shame, we decided to choose 5 items from the internal shame scale to measure shame in our research (Cook, 1988). These 5 items are "When I heard the critical comments, I do not think I meet the standards of the TJU basketball team", "When I heard the critical comments, I call into question of my physical strength and am disappointed in myself", "When I heard the critical comments, I started to doubt myself", "When I heard the critical comments, I feel embarrassed, sorry for teammates, and even sorry for school" and "When I heard the critical comments, I want to shrink myself and do not want others to see me" on 7-point scales(ranging from [1]=not at all and [7]=very much) to measure feeling of shame.

3.2.4. Measures of Feeling of Anger(AN)

In this research, anger comes from the situation which others' responsibility for the negative events but the sufferer is me (Averill, 1983). Four items we developed were "When I heard the critical comments, I feel I am angry", "When I heard the critical comments, I feel irritated", "When I heard the critical comments, I want to argue with them" and "When I heard the critical comments, I feel more angry than I am ashamed" on ranging from [1]=not at all and [7]=very much to measure anger.

3.2.5. Measures of Adaptive Consumption Behaviors (AC)and Compensatory Consumption Behaviors(CC)

We asked participants to read the concepts of adaptive consumption and compensatory consumption, "Adaptive consumption behavior—consumers purchase products or services for self-improving" and "Compensatory consumption behavior—consumers purchase products or services for self-repair" (Choi, Wang & Chen, 2018). Then in order to distinguish them, participants need to give choice answer between these two kinds of products (1=adaptive consumption, 2=compensatory consumption). Then, we designed each of 2 items to measure adaptive/compensatory consumption behaviors, "1. I would choose fitness center." "2. I would buy fitness center." "3. I would choose books."  "4. I would buy books."(1=not at all, 7=very much).

4. Empirical Investigation

4.1. Demographic Analysis Results

This research collected 353 questionnaires from Chinese college students by using 'WenJuanXing'. We deleted the questionnaires which participants failed to pay enough attention and 323 questionnaires were remained finally, which includes 107 of Type 1 (failure caused by my mistake) and 216 of Type 2 (failure caused by others‘ mistake). All the details are shown in Table 1.

Table 1: Demographic Analysis Results

4.2. Reliability and Validity

This research conducted principal component factor analysis to check the reliability and validity of the scales.

Construct reliability was measured by Cronbach‘s alpha (or α) to verify the internal consistency in SPSS 22.0 program. As shown in Table 2-1, Table 2-2, there were six principal components consisting of shame, anger, rightness of criticism to self (other member), rightness of criticism to social group (caused by self vs. caused by other member), compensatory/adaptive consumption behaviors and factor loading values of items indicated that all the items of each construct were judged to converge to their own construct. And except adaptive consumption concept, all other constructs‘ Cronbach‘s alpha was above 0.7.

Table 2-1: Results of Analyzing Components (My mistake)

Table 2-2: Results of Analyzing Components (Others‘ mistake)

4.3. Correlations among Constructs

Discriminant validity could be supported when average variances expected (AVE) are 0.5 or above and lager than the squared between-constructs correlation coefficients (Fornell and Larcker, 1981). The results of the analysis which include correlation coefficient, the squared correlation coefficients, AVE and composite reliability (CR) are shown in Table 3-1，Table 3-2. All the constructs except the adaptive consumption are discriminant from each other.

Table3-1: Results of Correlations Analysis (My mistake)

Note: the numbers of the diagonal (diagonal brackets) mean AVE (CR). And the numbers of the non-diagonal brackets mean the squared between-constructs correlation coefficients

Table3-2: Results of Correlations Analysis (Others‘ mistake)

Note: the numbers of the diagonal (diagonal brackets) mean AVE (CR). And the numbers of the non-diagonal brackets mean the squared between-constructs correlation coefficients

4.4. Testing Hypotheses

Objects of others' criticism after my in-group failed can be divided into my mistake and other‘s mistake, which are concerned with the differences between individual identity and social identity. We will focus on the differences of testing each hypothesis between the objects.

The fitness results in Table 4-1, and Table 4-2 showed the models met the acceptable level of goodness-of-fit.

Table4-1: Results of Testing Hypotheses (My mistake)

Note: Significance level(α)=.05

Table4-2: Results of Testing Hypotheses (Others‘ mistake)

Note: Significance level(α)=.05

The results of structural equation model analysis by AMOS 21.0 program were shown in Table 4-1 and Table 4- 2. When the failure caused by my mistake (Table 4-1), except H2-1 and H6, all hypotheses were accepted (p<0.05). In the failure caused by others‘ mistake (Table 4-2), except H3-1, H4-2 and H5, all hypotheses were accepted (p<0.05).

5. General Discussion

5.1. Research Summary

This study investigated whether consumers would feel ashamed or anger according to the rightness of observers‘ criticism when they experienced in-group failures caused by my/others‘ mistake, and also explored whether the shame (anger) positively influenced adaptive (compensatory) consumption behaviors.

Specifically, the results of this research are as follows

Firstly, emotions have an important role in the individuals‘ perception or even decision-making (DeSteno, Petty, Wegener & Rucker, 2000). Individuals who feel ashamed tend to choose adaptive consumption behaviors, and individuals who feel anger would choose the compensatory consumption behaviors. Whereas, in the failure caused by myself, shame positively affected adaptive consumption behaviors, and in the failure caused by others‘ mistake, anger positively affected compensatory consumption behaviors.

Secondly, when observers‘ criticism to the failure caused by my mistake is focused on me, perceiving the rightness would increase shame and decrease anger. While in the failure caused by others‘ mistake, the perceived rightness of criticism to the others only increased anger. It is easy to understand that most customers dislike to be criticized because of others‘ mistake.

Thirdly, when the object of criticism focused on their social in-group, and they were under the group failure caused by their mistake, the perceived rightness of the criticism would not decrease shame. Maybe when customer agrees with his/her duty to in-group failure, no matter whether the object of criticism is him/her or the in-group, customer would always feel more ashamed. However, individuals under the failure caused by others‘ mistake, the perceived rightness could not decrease their anger, maybe most customers dislike to be criticized because of others‘ mistake.

5.2. Theoretical Contributions and Managerial Implications

The results of this study demonstrated that observers‘ criticism to the different sources triggering the in-group failure could lead to different negative emotion, which push consumers to choose different type of consumption to regulate the negative emotion. In view of Belk's self-extension theory (1988) that consumers regard social groups as part of themselves, according to the object difference criticized (I, my social group, group member), customers would feel different main negative emotions, shame or anger, and choose different consumption behaviors, adaptive or compensatory. These findings will advance the theory related to the emotion-based choice.

Emotion science is undergoing revolutionary phases to influence the theory of consumers‘ decision making (Keltner & Lerner, 2010). Cognitive contents of eliciting emotion could lead to the place of feeling specific incidental emotion, which could influence consumption type. In the process of marketers‘ distributing their products or services to customers, it is necessary to consider how they approach their customer with the cognitive contents which induce the specific emotion, which in turn, could lead to distinct consumption type. Based on our findings we could make the following recommendations to marketers. Firstly, marketers should notice this interesting phenomenon that consumers may be inclined to choose one type of product/service and avoid the other type of product/service according to their negative emotional state. Marketers should notice dynamic changes in consumer sentiment and predict consumers' behavior. They should check bad emotions consumers recall and should develop the product messages appropriate to the emotion (Choi, Wang & Chen, 2018).

Secondly, marketers should note the different reactions of consumers between angry and ashamed emotions, and make efforts for consumers‘ avoiding explicit identity connection with the product/service that are not conducive to consumer emotional recovery (Rustagi & Shrum, 2018).

5.3. Limitation and Future Research

This study has some limitations that can be explored in future research.

First, we only compare the two emotions of shame and anger. However, the emotions of consumers are diverse and unpredictable, so follow-up research should provide more types of emotions caused by consumer failure. For example, behavioral effects of shame versus guilt on consumption type can be the research subject.

Second, our study targets Chinese college students, and the group characteristics are relatively single. To generalize the concept, the response to the failure could be explored more in view of cultural differences (Zhang, 2018; Ramana & Retnosari, 2018).