The social activist: conformity to the ingroup following rejection as a predictor of political participation

Abstract

Humans have an overwhelming desire to belong to social groups and if rejected, may act ingratiatory, and conform more easily, to a new accepting group (Williams, 2007) to establish themselves as a valued group member. We investigated whether following rejection, individuals were more willing to participate in and condone varying political actions against an excluding and opposing organization on behalf of a new organization. Results showed that rejected participants were more inclined to be politically active toward the outgroup, and this effect was magnified in those high in rejection sensitivity. Effects were mediated by thwarted inclusionary needs, suggesting that actions directed toward an outgroup may be driven by pro-social belongingness needs rather than outgroup aggression.

Keywords::

• conformity
• rejection
• political action
• inclusionary needs

In August 2011, riots broke out in London, spreading across the nation lasting for several days. Recently, a peaceful demonstration in a Stockholm suburb in Sweden was attacked by a group of neo-Nazis. The social and economic stresses of these neighborhoods are enormous. The explanations provided for these events seem to center on marginalization, social exclusion (Bäck, Bäck, & Garcia-Albacete, 2013; Topping, 2011), and the need to feel a sense of belongingness (Sarnecki, 2013). Sarnecki (2013) refers to the Nazis attacking the demonstration in Stockholm as striving for belongingness; a notion that receives support in personal accounts of former activists (Orr, 2013). This notion is intriguing as it suggests that action against another group might be driven by conformity to the ingroup in an attempt to secure a position within the own group. Even though the explanations mentioned may be based on research, there is to date no research that has systematically established a causal link between social rejection and political actions against an opposing and rejecting group.

Humans are innately social, striving for interactions with others and to be part of social groups. Consequently, individuals thrive on the opportunity to form meaningful relationships by joining various social groups (Baumeister & Leary, 1995). Recent research has considered the social benefits that political groups and participation provide (Bäck et al., 2013; Bäck, Teorell, & Westholm, 2011). A group fulfills an individual's social needs and if they are ostracized, rejected, or excluded from the group, this can have substantial impact on well-being, thwart social needs and elicit pain (Eisenberger, Lieberman, & Williams, 2003; Williams & Zadro, 2005). Following exclusion individuals are more likely to conform, seek acceptance, seek opportunities of social interaction and are often more aggressive (Williams, 2007). Thus, it is possible that rejected individuals may conform to group norms and engage in political activity to fortify thwarted needs of belongingness. In this paper, we focus on the effect of rejection on willingness to participate in political action against an opposing group.

We propose that following rejection, an individual will seek opportunities for social inclusion which may be provided by a political group. This is supported by recent research showing that some people participate in protests solely due to social aspects (Bäck et al., 2013). Hence, in this paper, we focus on the extent to which social factors, specifically rejection, may drive political actions against an opposing group.

Rejection and conformity as predictors of political activity

Rejection is the explicit statement that an individual or group is not wanted (Williams, 2007). Rejection, social exclusion and ostracism are usually analogously referred to as ostracism, even though they conceptually differ. Social exclusion is defined as an individual being kept apart from other groups, while ostracism is defined as being ignored and excluded (Williams, 2007). There is an ongoing debate about these conceptualizations, but there is no consensus regarding how they should be treated (Bernstein & Claypool, 2012a, 2012b; Smart Richman & Leary, 2009; Williams, 2009). In the present paper, we focus on rejection and will use the term rejection in relation to our hypotheses and results, but the literature review covers both exclusion and ostracism as well.

Four needs are fulfilled by membership in a group: the needs for belonging, self-esteem, control, and meaningfulness. The negative impact resulting from rejection is believed to be the consequence of these needs being thwarted. Often after the manipulation of rejection/exclusion, participants report strong feelings of anger, frustration, sadness, and lowered self-esteem, feelings of control, and a meaningful existence (Williams & Zadro, 2005).

To remove this distress, these needs have to be fortified and the extent to which each of the four needs is threatened mediate the extent to which an individual acts in a pro-social, ingratiatory manner, or in an aggressive, hostile manner. Specifically, if the need to belong or self-esteem (inclusionary needs cluster) are threatened, behavior is likely pro-social, such as conformity to group norms, whereas if feelings of control and meaningful existence (power-provocation needs cluster) are threatened, the response is hostility and aggression (Williams, 2007). Only a few studies have demonstrated the conforming effects of rejection, and even fewer the meditational effects of the thwarted inclusionary needs. We suggest that these are highly relevant when explaining actions against other groups. Even though such actions may appear to be aggressive, an alternative explanation is that they are actually motivated by concerns for belongingness to the ingroup.

Available research shows that following cyber-ostracism individuals are more susceptible to the Asch paradigm, conforming to clearly wrong answers provided by the group (Williams, Cheung, & Choi, 2000). In addition, ostracized individuals are more compliant (Carter-Sowell, Chen, & Williams, 2008) and obedient (Riva, Williams, Torstrick, & Montali, 2014) than included individuals. However, the ingratiatory, conforming actions only occur to a new group as it presents a realistic opportunity of inclusion, rather than the group that was the source of rejection (Maner, DeWall, Baumeister, & Schaller, 2007). Bernstein, Young, Brown, Sacco, and Claypool (2008) and Bernstein, Sacco, Brown, Young, and Claypool (2010) support this by showing that individuals recalling an episode of rejection, as compared to inclusion, were more accurate at detecting, and preferred, “genuine” smiles to “fake” smiles. Furthermore, meditational analyses indicated that this was partially mediated by the inclusionary needs, indicating that a threat to these needs means individuals become more adept to opportunities of inclusion (e.g., genuine smiles).

Consequently, rejection may impact participation in political activities, by providing opportunities for social interactions, and conforming to a new group's norms secures a position within that group minimizing the risk for future rejection. Thus, our first hypothesis is that rejection from one group will increase willingness to engage in and condone political action against that group on behalf of a new group compared to individuals who are not rejected (H1).

We here argue that actions against another group may be driven by belongingness needs even though they may appear aggressive. This notion is supported by research indicating that acting against an outgroup may in fact be driven by pro-social belongingness needs. For instance, qualitative research on gang cultures indicated that by actions against the outgroup an individual can climb on the hierarchical status ladder of the ingroup (Pinizzotto, Davis, & Miller, 2007; Stretesky & Pogrebin, 2007). Furthermore, conflicts between gangs provide a sense of belongingness to the own gang (Walker-Barnes & Mason, 2001). Hence, we expect that political actions against an opposing group following rejection is driven by inclusionary needs for belonging to the ingroup (H2).

Even though everybody reacts negatively to rejection, there are individual differences in one's sensitivity to rejection. Rejection sensitivity (RS) is a disposition to anxiously expect, readily perceive, and intensely react to, rejection, explaining why some individuals are more vulnerable to rejection than others (Berenson et al., 2009; Downey & Feldman, 1996; Downey, Khouri, & Feldman, 1997). Highly rejection sensitive individuals tend to conform more to group norms, if their identity is threatened (Romero-Canyas et al., 2010). If our argument that rejected individuals are more strongly motivated by social concerns to participate in actions against another group is valid, we should also see an effect of RS on these outcomes. Hence, we expect that rejected individuals who are highly rejection sensitive, will be more willing to engage in, and recommend, actions against another group compared to those low in RS (H3).

Thus, the present research aims to experimentally investigate whether rejection leads to increased willingness to engage in actions against an opposing, and rejecting, organization following acceptance into another organization. We also aim to assess the extent to which thwarted inclusionary needs mediate this relationship, and if individual levels of RS moderate effects.

Method

Participants

Forty students (20 male, 20 female, mean age = 23.89) were recruited at Lund University campus to a study investigating student views of student politics in exchange for a lottery ticket.

Design and procedure

A between-subjects design was used to assess whether there was an effect of rejection from a group on willingness to participate in, and condone political action against an opposing (and rejecting) organization. The independent variables were condition (non-rejected/rejected), and RS. The dependent variables were willingness to participate in, and condone, action against an opposing organization, and the extent to which the inclusionary and the power-provocation needs were thwarted.

On arrival, participants were given a demographic and RS questionnaire. Participants were then introduced to the proposal of implementing tuition fees at Swedish Universities. Previous research shows that most Swedish students (95%) oppose this (Bäck et al., 2013). Two participants were removed from the analyses because they supported tuition fees. Participants read a fictional newspaper article introducing them to the issue of tuition fees. When finished, they were asked to write a short statement detailing their beliefs about tuition fees. They were told this would be used as a proposal to join the (fictious) organization the International Students Union (ISU). Participants were not told the ISU's stance on the issue.

Participants were then informed that before joining the organization, their compatibility with it must first be checked by presenting their statement to a representative of the ISU, who also worked at the Department of Psychology, where the experiment was conducted. The experimenter said that she would just see whether the representative was available to look at the proposal and then left the room for five minutes. When the experimenter returned, the non-rejected participants were told that the ISU representative thought their beliefs were compatible with the organization and that they would like to gain some feedback from their new member regarding future political actions. To the rejected participants, the experimenter said that the representative thought their beliefs were not compatible with the ISU, and thus they could not be a part of the organization. The experimenter then said that she ran into a colleague on her way back to the lab, who also happened to be a representative of another student organization, the European Students Union (ESU). She informed the participant that this organization would be happy to have them as new members and explained that the ESU would like feedback on future political actions against an opposing organization (which was also the rejecting organization, the ISU). A few things should be noted here: firstly, the participants' allocation to conditions was purely randomized; and secondly, even though both the ISU and ESU opposed the implementation of the tuition fees, it was implied that the opposing organization (the ISU for the rejected participants and the ESU for the non-rejected participants) actually advocated tuition fees. Finally, as our participants were all university students, and the organizations were student organizations, it can be expected that they felt some identification with the organizations. Thus, even though they were not members of the ISU to start with, the student identification should have led them to experience rejection when informed that they were not welcome in that group.

Then, the participants were presented with a (fictional) article describing their organization's past action. This article was identical for both conditions except for the names of the organization, with the included participants being given an article about the ISU and the rejected participants being given an article about the ESU. This article described vandalism and riots in London against increased tuition fees as organized by the group that had included the participant (ESU/ISU). Thus, this article functioned to inform the participant of their group's norms, in order for them to have a standard against which they could conform.

When finished, participants completed a questionnaire assessing willingness to participate in, and recommendation of, action against the opposing group that they were informed “hindered their organization's aims.” This was to investigate whether the rejected participants were more willing to conform to group norms and/or act in an ingratiatory manner to the new and accepting organization by acting against the opposing organization. Finally, participants were presented with a questionnaire assessing the extent to which their four social needs were thwarted. All participants were debriefed and probed on whether they believed the cover story of the experiment. All participants were clueless as to the true purpose of the study, and all participants believed the representative(s) had checked their proposals.

Variable description

RS was measured using the short version of the RS scale (Downey & Feldman, 1996). Participants are asked to imagine themselves in eight different situations that describe things that people sometimes ask of others, for instance borrow money. For each situation, participants rate (a) how concerned or anxious they would be over the others' reactions (1 = very unconcerned to 6 = very concerned) and (b) to what extent they expect the others to help them in this situation (1 = very unlikely to 6 = very likely). To calculate a score of RS for each situation, the level of rejection concern (response to question a) is multiplied by the reverse of the level of acceptance expectancy (response to question b). An index of overall RS is then calculated by taking the mean of all scores for each situation, Cronbach's α = .85.

Willingness to participate in action against an opposing organization was measured using three items. The participants were advised that the conflicting organization opposed their beliefs and values and was hampering their efforts against preventing the implementation of the tuition fees. The participants were asked whether they would be willing to: “participate in [action/peaceful action/extreme action, e. g. sabotage] against the opposing group.” Responses ranged from 1 =  not at all willing, to 7 =  very willing (α = .86).

Recommended action against an opposing organization was measured with six items. The questions read: “do you recommend: sending a letter expressing our differing opinions/inviting them for a public debate/actively demonstrating outside their headquarters/sabotaging their efforts/vandalizing their headquarters. Answers ranged on a scale from 1 =  not recommended to 7 = highly recommended (α = .60).

Self-reported levels of needs. Nine items assessed the extent to which the four needs (belonging/self-esteem/control/meaningful existence) were thwarted during the experiment. This questionnaire was adapted from the needs-threat questionnaire used by Zadro, Williams, and Richardson (2004). Responses were assessed using nine-point Likert scales with varying anchors. Four items were combined to measure the inclusionary needs Cronbach's α = .86, the remaining five items were combined to provide a measure of the power-provocation needs, Cronbach's α = .55, which is low but sufficient. An average index composed of all nine items was also calculated and used as a manipulation check.1

Thus, the dependent variables were the indexes of willingness to take action against the opposing group, recommended action against the same group and the extent to which the inclusionary and power-provocation needs were thwarted. The independent variables were condition (rejection/non-rejection) and RS.

Results

Preliminary analyses and manipulation check

Because willingness to participate in action against another group and recommendation of actions against that same group could be considered as a higher-level factor of general action readiness against the group, we first performed a principal axis factoring with promax rotation to see whether the willingness and recommendation items would fall into one factor. As the variable to recommend sabotage had a communality >1, and hence prevented the analysis to reach a solution, it was removed from the analysis.2 The analysis revealed a two-factor solution, with two factors explaining 68% of the variance. Upon examination of the factor loadings, it was evident that the two peaceful recommended action items (sending a letter/ debate) were the only items that loaded low on the first factor and high on a second factor. Hence, we decided to drop these items and create an index of the rest of the willingness and recommendation items.3 Cronbach's α was high, .79.

Descriptives for all variables are presented in Table 1. To check that participants had been affected by the rejection manipulation, we performed a t-test with the independent variable condition (rejection/non-rejection) and the dependent variable the overall needs index.4 This indicated higher levels of thwarted needs among rejected (M = 4.02, SD = 0.83) compared with non-rejected (M = 2.93, SD = 1.18) participants, t(35) = − 2.42, p = .02, d = 1.07.

Table 1 Means, standard deviations, and correlations for all variables.

	M	SD	1	2
1. Activity index	3.36	0.94
2. Inclusionary needs	3.44	1.78	.53^*^*
3. Power-provocation needs	3.43	1.10	.22	.33^*

^*p < .05, ^**p < .01.

Main analyses

The first hypothesis was that rejected individuals should display higher levels of political activity against an opposing, and rejecting organization than non-rejected. An independent t-test, using condition (rejected/non-rejected) as the grouping factor and the political activity index as dependent variable, revealed a significant difference, t(36) = − 2.32, p = .026, d = 0.76. Rejected individuals rated themselves higher on political activity against the opposing group (M = 2.82, SD = 1.20) than the non-rejected (M = 2.02, SD = 0.88).5 Consequently, rejection affects the individual's action readiness against an opposing organization.

The second hypothesis stated that thwarted inclusionary needs would mediate the relation between rejection and political activity against the opposing organization. Thus, acts against the outgroup may be driven by inclusionary needs for belonging to the ingroup.

This was tested using the Preacher and Hayes’ (2008) macro for Statistical Package for the Social Sciences (SPSS) allowing multiple mediation tests of both inclusionary and power-provocation needs on the outcome variable simultaneously. This method also uses bootstrapping, which is especially recommended for small sample sizes (Preacher & Hayes, 2004). Condition (rejection/non-rejection) was entered as the independent variable, the inclusionary needs and the power-provocation needs as two separate mediators, and the political activity index as the outcome variable.

Mediation analyses based on 1000 bootstrapped samples using bias-corrected and accelerated 95% confidence intervals showed a significant effect of rejection on inclusionary needs such that when rejected, inclusionary needs became increasingly thwarted compared with those non-rejected, as seen in Figure 1. The analysis also revealed that rejection did not have an effect on the power-provocation needs cluster. There was a close to significant effect of the inclusionary needs on the outcome, while this was non-significant for the power-provocation needs. Furthermore, there was a significant total effect (path c) of rejection on the political activity index and a non-significant direct effect (path c) indicating full mediation.6

Figure 1 Mediation model for the mediation of the inclusionary and power-provocation need clusters on the dependent variable, where the solid lines represent significant associations.

Consequently, the results indicate that political activity against an opposing organization is motivated by inclusionary needs. This implies that intergroup conflict may, at least partly, be driven by conformity to the ingroup as a way to secure a position within a new group.

The third hypothesis concerned the moderating effect of RS on the effect of rejection on political activity. A hierarchical regression model with the single predictors condition (dummy-coded with non-rejected as reference category) and RS entered in Step 1, and the interaction between these in Step 2, with the political activity index7 as dependent variable, was performed. As shown in Table 2, there is a significant main effect of condition, confirming our earlier t-test. More importantly, there is an interaction effect of condition and RS on political activity, with those rejected and high in RS more likely to engage in political activity against another group, than if they are low in RS. These results confirm earlier findings that RS moderates the willingness to engage in political actions on behalf of ones group (Bäck et al., 2013). Furthermore, it supports the overall argument of the present research; that participation in actions against another group may be motivated by belongingness.

Table 2 Hierarchical regression analysis of rejection and RS on willingness to engage in political activity against another group.

		B	β	SE	p-Value	R^2
Model 1	Intercept	1.98		.48	< .001	.08
	Rejection^a	0.82	0.34	.39	.041
	RS	− 0.36	− 0.12	.27	.195
Model 2						.36
	Intercept	3.61		.57	< .001
	Rejection X RS	0.39	1.48	.10	< .001

a Rejection is a dummy variable coded as 0 if non-rejected, and 1 if rejected.

General discussion

Humans have a fundamental motivation to belong to groups (Baumeister & Leary, 1995; Williams, 2007). Recent research is beginning to consider the social benefits that political participation provide and has revealed the key role that these social benefits play in whether an individual becomes politically active (Bäck et al., 2013). When explaining action against other groups under the agenda of politics, it has been suggested that social exclusion, marginalization and belongingness may be key factors (Bäck et al., 2013; Orr, 2013; Sarnecki, 2013; Topping, 2011). However, to our knowledge, no prior studies have experimentally investigated this relationship.

Research on rejection reveals the massive distress, and social pain it causes (Eisenberger et al., 2003; Williams & Zadro, 2005) as a result of two fundamental need clusters (inclusionary and power-provocation) being thwarted. The extent to which these needs are threatened explain whether an individual acts in a pro-social or anti-social manner in an effort to fortify the thwarted needs (Williams, 2007). For example, rejected individuals may act pro-socially by conforming, seeking out social interactions and investing more heavily in a new and accepting social group if given the opportunity (Bernstein, Sacco, et al., 2010; Bernstein, Young, et al., 2008; Maner et al., 2007; Williams et al., 2000; Williams & Sommer, 1997).

We suggested that political groups may provide an opportunity to fulfill the inclusionary needs cluster and consequently, rejected individuals would be more willing to engage in political activity on behalf of a new group. In addition, we expected this willingness to be related to pro-social concerns for the ingroup. Overall, the results support this notion, indicating that rejected individuals are more willing to engage in political activity with their new group, than non-rejected ones, and that this is mediated by thwarted inclusionary needs. Moreover, we can see a moderating effect of RS on willingness to engage, such that there is a positive effect of RS for rejected participants on willingness to engage. This further supports the argument that concerns for belongingness are an important factor to take into account when explaining political activities against other groups. Hence, such conflicts may not necessarily be completely driven by ideological controversies between the groups but rather that actions directed toward an outgroup may be driven by pro-social belongingness concerns.

The results are in line with previous research demonstrating an increased receptiveness for social interaction following rejection (Pickett & Gardner, 2005). Furthermore, the results extend on earlier research showing that rejection increased participation in protest activities, such as demonstrations and petition signing (Bäck & Bäck, 2014).

Limitations

We did not find an effect of rejection on the power-provocation needs, which may warrant some concern. The lack of effect may be the result of the experimental paradigm, which is new and has not been previously tested. Participants were accepted into a new group within a short time frame following the rejection experience, and Williams state that in order for the power-provocation needs to be thwarted “re-inclusion with another individual or group appears unlikely” (Williams & Nida, 2011, p. 79). As a result, the lack of time without an opportunity for inclusion may explain the lack of effect on the power-provocation needs. In addition, the α for this cluster was relatively low (.55), indicating that besides the lack of time proceeding between the rejection episode and the inclusion opportunity, the questions may not have been properly adapted to the new paradigm. The questions used were adapted from the Cyberball paradigm and although we took care to adequately adapt the items, it is possible that the changes were insufficient for this new paradigm, as the two situations are very different. Hence, we are at this point unable to draw any strong conclusions with regard to the power-provocation needs cluster with respect to political engagement against another group. With that said, we did not expect a mediation effect of the power-provocation needs, but rather that the belongingness needs should be decisive for the outcome. Hence, even though methodologically problematic, the results are in line with our expectations. In addition, the moderating effect of RS may support the fact that it is actually pro-social concerns that are important in this situation.

Another drawback is the small sample size, which may make it difficult to detect significant differences. However, we still find significant effects indicating that they should be fairly strong. Still, the small sample size may also be one reason that we did not see any effects on the power-provocation needs, especially if the items were not appropriately adapted.

Finally, this article only reports results from one experiment, and replications are needed even though the results follow previous research on rejection and conformity. Hence, future research needs replicate the found effects using a larger sample.

Implications

Extending on our results, it could be the case that more violent political actions may be motivated by the desire to belong and a fear of rejection from the political group, rather than from aggression. Thus indicating that although these activists are acting in an anti-social manner to an opposing group, it may not be solely driven by anti-social needs, but rather by a desire for acceptance. This is in line with qualitative reports where gang members display outgroup aggression to prove ingroup loyalty (Pinizzotto et al., 2007). Results are also in line with previous research that has demonstrated an increased desire for social interaction, increased conformity, and an increased desire to act in an ingratiatory manner to a new and accepting group (Bourgeois & Leary, 2001; Maner et al., 2007; Williams et al., 2000; Williams & Sommer, 1997).

Ostracism research is beginning to consider societal level and intercultural exclusion, and this is likely to have large implications in relation to the present research (Aydin, Fischer, & Frey, 2010; Williams, 2007). If one group is continuously excluded within a society, such as marginalized groups can be considered to be, this may have implications regarding the political activities they engage in. An increasing multicultural and diverse society is likely to be enhancing these issues and if individuals are forced to assimilate to mainstream values, they may feel excluded from society if they do not integrate.

Concluding remarks

This research has provided a novel and promising explanation for why individuals may engage in political actions against a group. To sum up, our results point to the human social need to feel belongingness and political groups as providing such opportunities, when explaining political activity. Thus, political action against some groups may be driven by conformity to ingroup norms, rather than actually wanting to hurt the outgroup. This does not mean that we should condone the actions of such individuals if they cause pain and suffering to other people, but it contributes to our understanding of why such acts may be performed. We would here also like to stress that we do not believe that only social needs motivate individuals to become active—also other more instrumental incentives, where individuals are driven by ideology and the goal of influencing collective outcomes in a certain direction, as suggested by both Klandermans (Klandermans & van Stekelenburg, 2013) and the collective action model (Bäck et al., 2011; Olson, 1965), are of course also important to take into account. We do, however, believe that it is important to consider the mechanisms of rejection with regard to political action and that this investigation has highlighted an area that is lacking in research—hopefully it can create the necessary platform for future research on such topics. Especially, there is a need for research that combines ideological incentives and social incentives presented in the present research.

Additional information

Funding

This work was supported by The Swedish Research Council [grant number 421-2011-1333].

Notes

^1. The inclusionary needs consist of the items “I did not feel accepted by the group” and “I felt left out when I heard the group's statement” (Need for Belonging); “I felt that the group did not consider me as a valuable and pleasant person” and “I felt I did not quite fit into the group” (Need for Self-Esteem), while the power-provocation needs consist of the items: “I felt frustrated when I heard the group's statement” and “I felt I had control of the situation” (Need for Control); “I felt that what I did [e.g., write about myself] had some impact on the group's decisions,” “I felt as if I was not there for the group” and “I felt as if my life was meaningless” (Need for a Meaningful Existence).

^2. This is referred to as an ultra-Heywood case.

^3. We also ran all analyses with both the separate indices for willingness (action/peaceful action/extreme action items), and recommendation (demonstrate/sabotage/vandalize) as well as, for an index of the two peaceful recommended action items, separately.

^4. There was a significant difference between the rejected and non-rejected participants on the inclusionary needs, t(36) = − 3.96, p < .001, d = 1.29, such that rejected participants reported higher levels of distress (M = 4.41, SD = 1.03, and M = 2.47, SD = 1.86, for the rejected and non-rejected participants, respectively). This difference was not significant for the power-provocation needs, t(36) = 0.10, p = .92, d = 1.08.

^5.t-Tests were significant, or close to significant for the recommendation index, t(36) = − 2.57, p = .14, d = 0.84, the peaceful recommendation index, t(36) = − 1.70, p = .098, d = 0.75, and for the willingness index, t(36) = − 1.89, p = .066, d = 0.62, such that rejected individuals were more willing to recommend actions (M = 2.53, SD = 1.16), peaceful actions (M = 6.44, SD = 0.76), and participate in actions (M = 3.11, SD = 1.34) than non-rejected participants (M's = 1.72, 5.89, and 2.33, SD's = 0.73, 1.20, and 1.17, respectively).

^6. The analysis for the recommended action index showed similar results (a¹: B = 1.93, p < .001, a²: B = − 0.04, p = .92, b¹: B = 0.17, p = .13, b²: B = 0.19, p = .22, c: B = 0.81, p = .01, c′: B = 0.48, p = .196), as did the analysis for the recommended peaceful actions (a¹: B = 1.93, p < .001, a²: B = − 0.04, p = .92, b¹: B = 0.28, p = .02, b²: B = − 0.21, p = .20, c: B = 0.55, p = .09, c′: B = 0.01, p = .98), and that for the willingness index (a¹: B = 1.93, p < .001, a²: B = − 0.04, p = .92, b¹: B = 0.22, p = .15, b²: B = 0.15, p = .45, c: B = 0.77, p = .066, c′: B = 0.36, p = .47), although somewhat less strong results, the pattern is the same across all dependent variables.

^7. The interaction was significant for the recommendation index, B = 0.26, SE = .09, p = .004, for the willingness index, B = 0.41, SE = .11, p < .001, and for the recommended peaceful actions, B = 0.33, SE = .08, p < .001.