the effects of social identity threat

The following ad supports maintaining our C.E.E.O.L. service

The Effects of Social Identity Threat on Reactions to Normative andDeviant Group Members: A Subjective Group Dynamics Approach

«The Effects of Social Identity Threat on Reactions to Normative and Deviant GroupMembers: A Subjective Group Dynamics Approach»

by R. Isabel Pinto; M. Jose Marques; Frederico Guilherme

Source:Social Psychology (Psihologia socială), issue: 30 / 2012, pages: 1932, on www.ceeol.com.

http://www.ceeol.com/

http://www.dibido.eu/bookdetails.aspx?bookID=d1e91a38-03f5-4166-a958-30489ae8fdfd

http://www.ceeol.com

Frederico Guilherme, Isabel R. Pinto, José M. Marques

The Effects of Social Identity Threat on Reactions to Normative and Deviant Group Members: A Subjective Group Dynamics Approach1

Abstract: The present research draws from the postulates of subjective group dynamics theory (SGDT) (Marques, Abrams, Paez, & Hogg, 2001; Marques & Paez, 2008) and previous research on the black sheep effect (Marques, Abrams, & Serodio, 2001; Marques & Yzerbyt, 1988). We found support to the SGDT hypothesis that the extreme differentiation between normative and deviant in‑group members is determined by the motivation to guarantee the in‑group’s positive distinctiveness. In our study, participants derogated deviant in‑group members and upgraded normative in‑group members in comparison to respectively similar out‑group members, when the in‑group’s positive distinctiveness was undermined by an unfavorable inter‑group comparison. In contrast, when participants were ensured about the in‑group’s positive distinctiveness, they favored both in‑group members in comparison to the out‑group targets. These results are consistent with SGDT, and are at odds with an explanation of extreme intra‑group differentiation as a result of the violation of positive in‑group expectancies (cf. Biernat, Vescio, & Billings, 1999).

Key words: black sheep effect, deviance, social identity, social influence, subjective group dynamics

Guy Fawkes was an English Catholic involved in the failed attempt to assassinate King James I in 1605. Marshall Henri Philippe Pétain was the Head of State of France who collaborated with the Nazi regime during World War II. Mordechai Vanunu is a former Israeli nuclear technician who in 1986 publicly revealed details of his country’s nuclear weapons program. These are illustrations of famous persons who were convicted for high treason. They were vilified and despised in their countries. They were also sentenced to severe penalties, including life imprisonment and death.

In order to understand why individuals despise and antagonize in‑group traitors, we must understand the reasons why they reject in‑group deviant members in general. By “deviant” we mean a member of a social group who behaves against the group’s valued norms (Marques, Abrams, Paez, Hogg, 2001; Marques & Paez, 2008). Treason is a particular case of deviance as it goes against a general norm of loyalty to the group. In

1. This research is part of Frederico Guilherme’s PhD project supervised by José M. Marques and supported by the Fundação para a Ciência e a Tecnologia (FCT), Portugal, Grant: SFRH/BD/46518/2008. José Marques is also an Associate Researcher at the Institute of Social Sciences, University of Lisbon. Correspondence concerning this article should be addressed to: Frederico Guilherme, Faculdade de Psicologia e de Ciências da Educação da Universidade do Porto, Rua Alfredo Allen, 4200‑135 Porto, Portugal, e‑mail: frederico‑[email protected].

Access via CEEOL NL Germany

20 FREDERICO GUILHERME, ISABEL R. PINTO, JOSÉ M. MARQUES

contrast, normative group members are those who behave accordingly with the group’s norms and, as such, they generally stand out less than the deviants.

According to subjective group dynamics theory (Marques, 2010; Marques, Abrams, Paez, et al., 2001; Marques & Paez, 2008), hostility toward deviant in‑group members is illustrated by the evaluative differentiation process coined the black sheep effect, by which individuals upgrade normative in‑group members and downgrade deviant in‑group members as compared to similar out‑group members (e.g., Marques & Paez, 1994; Marques, 2010; Marques & Yzerbyt, 1988; Marques, Yzerbyt, & Leyens, 1988). SGDT accounts for the black sheep effect as an outcome of people’s motivation to ensure that fellow in‑group members are consonant with the prescriptive standards that legitimate their beliefs on the in‑group’s positive distinctiveness (Marques, 2010; Marques, Abrams, & Serodio, 2001; Marques & Paez, 2008).

Indeed, previous research on subjective group dynamics has showed that group members are generally favorable toward normative in‑group members, while being more hostile towards deviant in‑group members than deviant out‑group members, when it is necessary to achieve or to restore the in‑group’s positive distinctiveness (e.g., Marques, Abrams, & Serodio, 2001). Our goal with the present study is to follow up that research in support of the idea that individuals favor normative in‑group members and derogate deviant in‑group members, more than similar out‑group members, when the in‑group’s positive distinctiveness is undermined. We suggest this is a way to achieve and sustain a positive social identity. In addition, we analyze how the motivation for a positive social identity can determine the advocacy of strategies of social influence, either socializing or punitive, toward in‑group deviants (cf. Pinto, Marques, Levine, & Abrams, 2010).

Subjective group dynamics and the black sheep effect

Subjective group dynamics theory (SGDT) draws from the general postulates of the social identification approach, including social identity theory (e.g., Tajfel, 1978; Tajfel & Turner, 1986) and self‑categorization theory (e.g., Turner, Hogg, Oakes, Reicher, & Wetherell, 1987) to which it adds a perspective about the role of deviance in structuring social relations (cf. Marques, Abrams, Paez, et al., 2001; Marques, Abrams, & Serodio, 2001; Marques & Paez, 2008). According to the social identification approach, individuals strive to achieve and to sustain a positive social identity by achieving and maintaining a positive differentiation as compared to relevant out‑groups in specific social situations (Tajfel, 1978). Namely, individuals assimilate themselves to an in‑group prototype and, as a result, develop in‑group favoring attitudes, which materialize in terms of biased judgments and discriminatory behavior regarding out‑groups and their members (Tajfel, Billig, Bundy, & Flament, 1971; Tajfel & Turner, 1986; Turner, et al., 1987).

SGDT, on its side, aims to extend the social identity approach to the analysis of intra‑group differentiation processes that occur in inter‑group contexts. Based on research on the black sheep effect, SGDT sustains that intra‑group differentiation may allow individ‑uals to subjectively validate the in‑group, namely by reinforcing their commitment to the norms that legitimate their beliefs on the in‑group’s positive value as compared to an out‑group (e.g., Marques, 2010; Marques, Abrams, & Serodio, 2001; Marques & Paez, 2008; Pinto, et al., 2010).

Since the initial work by Marques and colleagues (Marques & Yzerbyt, 1988; Marques, et al., 1988), the black sheep effect was found in a wide variety of social contexts, under

21THE EFFECTS OF SOCIAL IDENTITY THREAT ON REACTIONS TO NORMATIVE...

a range of different experimental conditions (e.g., Biernat, et al., 1999; Branscombe, Wann, Noel, & Coleman, 1993; Breakwell, Vignoles, & Robertson, 2003; Eidelman & Biernat, 2003; Hichy, Mari, & Capozza, 2008; Jetten, Summerville, Hornsey, & Mewse, 2005; Khan & Lambert, 1998; Lewis & Sherman, 2010; Nesdale & Brown, 2004). In this line of research, some authors proposed alternative accounts for the occurrence of the black sheep effect. In particular, Biernat, Vescio, and Billings (1999) relate the downgrading of in‑group deviant members to the aversive emotions triggered by the violation of higher positive expectancies toward the in‑group than the out‑group by deviant in‑group members. Below, we focus on the SGDT’s account on the black sheep effect. In a follow‑up section, we analyze Biernat, Vescio, and Billings’ (1999) research and alternative explanation.

Descriptive focus and prescriptive focus in judgments of in‑group members

SGDT suggests that judgments of in‑group members ensue from the articulation of a descriptive and a prescriptive focus on the perception of individuals’ behavior (Marques, 2010; Marques, Abrams, Paez, et al., 2001; Marques & Paez, 2008). Descriptive focus emerges when perception is driven by the accentuation of inter‑category differences and intra‑category similarities. In such cognitive contexts, individuals simultaneously concentrate on perceived similarities between instances of the same category and on perceived differences between instances of contrasting categories (Haslam & Turner, 1992; Oakes, Haslam, & Turner, 1994; Turner, et al., 1987). Social groups thus become cognitively construed in terms of those attributes that, in each particular social context, best represent group prototypes (Turner, et al., 1987), and individuals (including the self) come to be perceived as in‑group or out‑group members (Marques, Paez, & Abrams, 1998).

According to the social identification approach, group prototypes correspond to the groups’ contrasted positions along dimensions that SGDT defines as descriptive norms. Examples of descriptive norms may include skin color when ethnic differentiation between groups is relevant, or particular chants voiced by soccer fans in support for their respective teams during a match (Marques, Abrams, & Serodio, 2001; Marques, et al., 1998; Pinto, et al., 2010). The position occupied by any individual on a descriptive dimension defines that individual’s membership to one of the salient groups in those particular contexts.

However, there can be situations in which, after inter‑group distinctiveness is cognitively established through descriptive focus, individuals’ attention is directed at group members’ behavior on dimensions that are not relevant for inter‑group differentiation and category inclusion, but that define the value ascribed to the groups and their members. Examples of such dimensions, which SGDT designates as prescriptive norms, include “honesty”, “fair‑play”, “solidarity”. Prescriptive norms stand for moral conventions (cf. Cialdini, Reno, & Kallgren, 1990; Reno, Cialdini, & Kallgren, 1993) that are common to every social group and thus do not allow for inter‑group differentiation. However, they help define the value assigned to the group and hence the perceived legitimacy, or subjective validity (Pinto, et al., 2010), of group members’ beliefs on the positive value they ascribe to their group.

SGDT suggests that deviant behavior triggers a prescriptive focus on the part of the observers, who will devote their attention to the extent to which the actor contributes positively or negatively to the overall value assigned to their group. When the actor is an in‑group member, deviant behavior jeopardizes belief on the relative superiority of the in‑group. As a result, this member should be strongly derogated, as a means to legitimize


the in‑group’s positive identity in comparison to the out‑group. This process leads to extreme evaluative appraisals of deviant and normative in‑group members, who are respectively threatening and supportive to the in‑group’s positive distinctiveness, in comparison to out‑group similar targets, i.e. the black sheep effect.

Identity threat and the black sheep effect

In line with the above reasoning, some research suggests that the black sheep effect is more likely to occur when the in‑group’s image is undermined and it seems necessary to restore its positive stand (e.g., Marques & Paez, 1994; Marques, Abrams, & Serodio, 2001). For example, Marques, Abrams, and Serodio (2001, Experiment 3) categorized participants in two groups based on a bogus imagination test. Some participants were then informed that their in‑group type of imagination was undoubtedly superior to the other (a secure positive identity scenario), whereas other participants were told that it was not sure which type of imagination was superior, and that the purpose of the study was to clarify this issue (an insecure positive identity scenario). All participants were then presented with four group members who adopted a normative ethical opinion and one member who adopted a deviant opinion.

Results showed that only in the insecure positive identity scenario, participants judged the normative in‑group member more favorably than the normative out‑group member and derogated the deviant in‑group member in comparison to the deviant out‑group member. In the secure positive identity scenario, participants favored both in‑group members in comparison to the out‑group members. These results thus illustrate the idea that group members are more inclined to reject deviant in‑group members and uphold normative in‑group members when their identity’s positive value is undermined than when it is secure by comparison with a relevant out‑group.

Expectancy violation and the black sheep effect

Biernat, Vescio and Billings (1999) proposed an account of the black sheep effect based on a theoretical rationale alternative to subjective group dynamics theory. According to these authors, the derogation of in‑group deviants ensues from a state of negative mood arising from a perceived violation of positive in‑group expectancies. To test this hypothesis, Biernat, Vescio and Billings (1999) had White female university students playing a team game with a White or Black female confederate. Participants were told that the better their team result was, the more chances they had in winning a prize. Confederates delivered a high or low quality performance. After the game, participants were asked to judge their partner’s performance, competence and warmth and whether their partner’s performance had violated their expectancies. Before and after the game, participants were also asked to indicate the extent to which they were experiencing each of a series of positive and negative emotions. According to the results, low‑performing White partners elicited more expectancy violation and more negative mood change, in comparison to both the low‑performing Black and the high‑performing White partners. Furthermore, mediation analysis confirmed that perceived expectancy violation led to negative mood, which in turn led to more negative evaluations.

In our view, it seems difficult to support the expectancy violation hypothesis relying on the Biernat, Vescio and Billings’ (1999) study. Because this study included only White participants, there is a possibility that the significant correlation between expectancy


violation and negative mood was not related to social identity concerns per se, but to race‑based stereotypes. Participants might expect the White confederate to be a competent partner, not because she was a team member, but because she was White (cf. Whitehead, Smith, & Eichhorn, 1982). Thus, a study including also Black participants would be advised to account for this alternative possibility. Also, and perhaps more importantly, the expectancy violation measure was collected after participants interacted with the partner and after having answered to the mood and attitudinal scales. Thus, the question of causality remains open. Biernat, Vescio and Billings (1999) propose that the perception of expectancy violation leads to more negative moods and more negative evaluations of in‑group deviant members. However, according to the temporal ordering of the variables in their study, it is possible that the opposite causal order may be right.

The expectancy violation hypothesis is a compelling explanation for the black sheep effect and can account for some instances of social discrimination, perhaps not necessarily against in‑group deviants, but toward any in‑group or out‑group members who fall short of group expectations. Thus, we believe this hypothesis deserves further scrutiny and should be studied as an alternative explanation to the in‑group’s positive distinctiveness hypothesis supported by subjective group dynamics theory and research (cf. also Pinto, et al., 2010).

Advocated strategies toward in‑group deviant members

More recent research on SGDT has showed that, in association with evaluations of normative and deviant in‑group members, individuals advocate harsher social control strategies toward deviant in‑group members, and that this is associated with their motivation to defend the in‑group’s identity. For example, Pinto, et al. (2010, Study 3) informed students from two courses that they would organize a debate between students issuing from both courses. Participants were presented with in‑group or out‑group course normative and deviant targets as potential debaters. Based on Levine and Moreland’s (e.g., 1994) group socialization model, these targets were presented either as new to their course (new members), or as well‑established students (full members), or as students who wished to leave their course (marginal members). Participants were asked to evaluate a target who advocated a normative opinion and a target with a deviant opinion about the topic for the debate, and to advocate a series of socializing and punishing strategies as to deal with the target whom they agreed the least (all participants chose the deviant).

Pinto and colleagues (2010) found that the normative in‑group full member was the most favorably evaluated of all targets, and the deviant in‑group full member was the most unfavorably evaluated of all targets. In addition, the more negatively participants evaluated the deviant in‑group full member, the more they advocated a punishing reaction toward this target, whereas the more negatively participants evaluated the deviant in‑group new member, the more they advocated a socializing reaction toward this target. These findings suggest that the appraisals of deviant in‑group targets can be related to more socializing or punitive strategies, as a function of the level of threat the targets pose. More specifically, because in‑group full members are held responsible for upholding the group’s values and image (Levine & Moreland, 1994), any deviant behavior on their account poses a strong threat to the in‑group’s image. Thus, the group should be highly motivated to punish in‑group full members’ deviant behaviors. In contrast, deviant in‑group new members pose less of a threat to the group’s image, and thus trigger less punishing and allow more socializing reactions. In the present study, we follow this research by analyzing how punishing and


socializing strategies of social influence toward in‑group deviants can lead to the validation of a positive in‑group image.

Overview and hypotheses

We test reactions to group members in an inter‑group context which is either favorable or unfavorable to the in‑group, namely a situation where participants are informed about the in‑group’s positive value and a situation where they are informed of the in‑group’s negative value to an out‑group. This methodology allows us to test the following competing hypothesis, either (1) that the extreme reactions toward deviant and normative in‑group targets, which are characteristic of the black sheep effect, are triggered in the presence of undermining information to the in‑group’s image to a relevant out‑group (cf. Marques, Abrams, & Serodio, 2001), or (2) that the black sheep effect is triggered when in‑group deviant members contradict the in‑group’s higher standards in comparison to an out‑group (cf. Biernat, et al., 1999).

We presented two inter‑group settings to students of Psychology and Educational Sciences courses. In one setting, the in‑group (Psychology or Educational Sciences) course had more deviant members in comparison to the other course (Threatening setting), whereas in the other setting (Secure setting), the in‑group course had less deviant students. We asked participants to judge a deviant and a normative target, who were either in‑group course, or out‑group course students. In line with previous research (e.g., Marques, Abrams & Serodio, 2001), we expected the black sheep effect to emerge from the threatening, but not from the secure setting, i.e. participants would derogate the in‑group deviant member as compared to the out‑group deviant member, and would be more favorable to the in‑group normative member than to the out‑group normative member.

In addition, we asked participants to indicate their degree of support for a socializing and a punishing strategy aiming to reduce deviance in the in‑group. Previous research on subjective group dynamics theory confirms that the support for socializing and punishing strategies towards in‑group deviants is related to the implications of the deviant’s behavior for positive group identity (Pinto, et al., 2010, Study 3). Individuals can give more support toward a socializing reaction aimed at convincing in‑group deviants to change their views, if the in‑group deviant members pose little threat to the in‑group’s image (e.g., because they are new members); and adhere more to a punishing reaction, as to remind in‑group deviants of the social costs of their deviance, if the in‑group deviant members pose a strong threat to the in‑group’s image (e.g., because they are full‑time members). Either way, the support for the influencing strategies toward in‑group deviants should help individuals to assure a more positive view of the in‑group. In our study, we tested this hypothesis with an exploratory analysis on how the support for socializing and punishing collective strategies aiming to reduce in‑group deviance can lead to a more positive view of the in‑group.

Method

Participants and design

Participants were 24 Psychology and 26 Educational Sciences students attending the Univer‑sity of Porto, Portugal (80% female, M = 22.12 years, DP = 6.88). A 2 (Inter‑ group Setting: Threatening vs Secure) × 2 (Target’s Group: In‑group vs Out‑group) × 2 (Target’s Stance:


Normative vs Deviant) mixed design was used. Group Comparison and Target’s Group were between‑participants factors, and Target’s Stance was a within‑participants factor1.

Procedure

Participants were approached in public places in the Faculty building and asked to answer a survey about prejudiced attitudes and discrimination among university students. Participants were informed that a preliminary survey had measured Psychology and Educational Sciences students’ opinions about several minority groups in Portugal. This survey had, supposedly, highlighted the percentage of students in both courses who supported a discriminatory policy against Gypsies, namely “exiling Gypsies who are condemned for criminal activities”. Agreement with this policy had been previously shown to be infrequent and considered as undesirable by a different sample of individuals from the same population2. According to the Inter‑group Setting manipulation, participants either read that the in‑group course had proportionally more, or less students, who supported the discriminatory policy than the other course (18% vs 3%, or 18% vs 33%, respectively).

Before being presented with the information about the percentages of group members, participants answered to a series of social identification questions. Next, they were asked to evaluate two students who had given an opinion about the matter. The first target student supported the discriminatory policy, and declared that Gypsies are “untrustworthy and prone to engage in criminal behavior” (this statement was pre‑tested to be undesirable by another sample of individuals from the same population3). The second target student opposed the

1. Participant’s course was similarly distributed across conditions χ2(3, N = 50) < 1. Participant’s sex and age were unbalanced across conditions, χ2(3, N = 50) = 7.55, p = .056 and F(1, 46) = 4.020, p = .051, respectively. Apart from the change in their respective magnitudes, the analysis involving participants’ course, sex and age as covariates in the design of the experiment was similar to the analysis reported in the results section. We thus disregarded Participant’s course, age and sex in the analysis.

2. We asked 16 Psychology and 12 Educational Sciences students about their opinion on five anti‑Gypsies discriminatory policies, one of them being “exiling Gypsies who are condemned for criminal activities”. We asked participants to estimate the percentage of students within their course and Faculty that would support such a policy. Results indicated that students regarded the support for that policy as infrequent within their own course (M = 18.14%, SD = 17.71) and Faculty (M = 26.50%, SD = 21.57). Both results were significantly below a 50% threshold, t(27) = 9.52, p < .001 and t(27) = 5.76, p < .001, respectively. Also, we asked participants about their personal support for the policy on a 7‑point scale (1 = I totally disagree, 7 = I totally agree). Participants did not support the discriminatory policy (M = 2.34%, SD = 1.28, significantly below the middle point of the scale, t27 = 6.97, p < .001). There were no significant differences between Psychology and Educational Sciences students for the estimation of the frequency of support within the Faculty and personal support, t(26) always < 1.01, ns. But Educational Sciences students tended to regard support for the policy as more infrequent within their course (M = 10.92%, SD = 11.76) in comparison to Psychology students (M = 23.56%, SD = 19.76, t26 = 1.97, p = .060). Both these results were significantly below 50%, t(15) = 5.35, p < .001 and t(11) = 11.51, p < .001, respectively.

3. We asked 11 Psychology and 13 Educational Sciences students about their opinion on four statements of different university students whose courses were unidentified. They were asked to evaluate each statement, on a scale between 1 = “negative” and 7 = “positive”. We retained two statements for the main study: One statement that was evaluated favorably (M = 5.46%, SD = 1.02, significantly above the middle of the scale, t23 = 7.00, p < .001) and another that was evaluated unfavorably (M = 2.29%, SD = 1.04, significantly below the middle of the scale, t23 = 8.03, p < .001). We submitted these scores to a Participant’s Course (Psychology and Educational Sciences) × Statement (Favorable and Unfavorable) analysis of variance, with the latter measure as a within‑subject factor. Participants favored one statement over the other, as shown by the significant Statement main effect, F(1, 22) = 77.31, p < .001. Participant’s Course and Course × Statement yielded no significant effects, both F(1, 22) < 1. A further t‑test conducted on the two scores, after reversing the items for the


discriminatory policy, by noting that “any discriminatory judicial policies can only lead to more marginality and exclusion” and stressing the values of a “multi‑cultural, democratic and cohesive society” (pre‑tested as desirable). Both the deviant and the normative targets were presented as attending either the same course as the participant (in‑group) or the other course (out‑group). Participants were then asked whether they would support two civic groups formed by students of the same course as theirs. Both groups aimed to reduce prejudice among students of their course, but whereas one group advocated a persuasive strategy, the other advocated a punishing strategy to be applied to prejudiced in‑group students. In the end, participants answered to a second set of social identification questions.

Dependent Measures

Social identification. Participants indicated their degree of identification with their course at two different times. First, after being informed of the goals of the study and before the experimental manipulation, participants answered the following items: “Do you like the students of your course?” (1 = I don’t like them at all; 7 = I like them a lot), “To be a student of your course is an important part of your identity?”, “Do you trust the students of your course?”, “Do you treasure the friendship of the students of your course?”, “Do you believe students of your course are among the most qualified at the University of Porto?”, “Do you think about yourself frequently as a student of your course?”, “Do you believe students of your course are among the most altruists at the University of Porto?” (1 = not at all; 7 = totally). Second, at the end of the survey: “Do you think you have similar attitudes and values with the generality of the students of your course?”, “Do you believe the students of your course have high levels of morality?”, “Is it important to you to be seen as a student of your course?” (1 = not at all; 7= totally), and “What would you feel if you heard someone badmouthing students of your course?” (1 = very well; 7 = very bad). We averaged the answers to these items to a pre‑Identification and a post‑identification scores (Cronbach´s α = .72, and .67, respectively). This scores were significantly correlated, r = .634, p < .001.

Evaluation of targets. Participants were asked to evaluate each of the two target‑‑students, on the following item: “What is your impression of this student?” (1 = „negative”, 7 = „positive”).

Advocated strategies of social influence. This measure was inspired by the Political Action Intentions Index (cf. Iyer, Schmader, & Lickel, 2007). Participants were presented with descriptions of two hypothetical civic groups composed by students of the same course as theirs, each one advocating specific strategies aimed at reducing prejudice in their course. One group advocated persuasive strategies (“informative measures, like conferences, debates, cultural events”), whereas the other group advocated punishing policies (“disciplinary measures, like sanctions, fines, punishments”) to be applied to prejudiced in‑group students. Participants were asked to indicate their willingness to support each group in four ways, ranging from less to more costly: “joining the group’s mailing list”, “signing the group’s manifesto”, “joining the group”, and “recruiting other students to join the group” (1 = not at all; 7 = totally). Thus, two sets of measures were averaged to respectively assess participants’ willingness to participate in actions to support the persuasive group (four items, α = .88) and to support the punishing group (four items, α = .96).

unfavorable score (1 = 7; 2 = 6; … 6 = 2, 7 = 1), was non‑significant (t23 = 1.10, ns), suggesting that the favorable statement was as better evaluated as the unfavorable one was depreciated.


Results

Pre‑identification

To check for participants’ identification with the in‑group course before experimental manipulations, we ran an Inter‑group Setting (Threatening vs Secure) × 2 (Target’s Group: In‑group vs Out‑group) analysis of variance on the pre‑identification score. There was a significant difference in this measure between Targets’ Group conditions (In‑group: M = 4.69%, SD = .68 vs Out‑group: M = 5.45%, SD = .39, F(1, 46) = 22.99, p < .001, η2 = .33). This difference can only be attributed to chance, as the score was taken before participants received any information related to the experimental manipulations.

Evaluations

We expected the black sheep effect to occur more strongly in the unfavorable inter‑group comparison than in the favorable inter‑group comparison condition. To check for this prediction, we conducted an analysis of covariance (ANCOVA) Inter‑group Setting × Targets’ Group × Target’s Stance on evaluation of normative and deviant members, while controlling for the effects of pre‑identification.

The ANCOVA showed a significant Inter‑group Setting × Targets’ Group × Target’s Stance interaction, which was consistent with our main hypothesis, F(1, 46) = 5.51, p = .023, η2 = .111 (see Figure 1). We decomposed this interaction according to the Inter‑group Setting condition. The Targets’ Group × Target’s Stance interaction was marginally significant in the threatening setting and was not significant in the secure setting, F(1, 47) = 3.55, p = .066, η2 = .07 and F(1, 47) = 2.20, ns, respectively.

Figure 1. Evaluations of normative and deviant targets as a function of inter‑group comparison (unfavorable or favorable) and targets’ group (in‑group or out‑group).

1. Pre‑identification as a covariate had a reliable effect, β = .43, t = 2.65, p = .011.


In the threatening setting, we found a pattern which was consistent with the black sheep effect. Participants tended to evaluate more negatively the deviant in‑group member (M = 1.33%; DP = .49) than the deviant out‑group member (M = 1.92%, DP = .86; F1, 46 = 3.92, p = .094, η2 = .06) and evaluated more positively the normative in‑group member (M = 5.42%; DP = 1.08) than the normative out‑group member (M = 4.85%, DP = .90; F1, 46 = 4.62, p = .037, η2 = .09).

In the supportive setting, there was a Targets’ Group main effect, F(1, 46) = 6.56, p = .014, η2 = .12. An exploratory analysis of the simple effects suggested that the Targets’ Group effect was mainly due to the deviant target’s evaluation, which was higher toward the in‑group deviant (M = 2.50%, DP = 1.02) than toward the out‑group deviant (M = 1.64%, DP = .81, F1, 46 = 3.92, p = .054, η2 = .08). There was no significant difference between the evaluations of the normative in‑group (M = 5.50%, DP = 1.29) and normative out‑group members (M = 5.55%, DP = 1.51, F1, 46 = 2.59, p = .11, η2 = .05).

Evaluative judgments, social control strategies, and identity motivations

We expected that the derogation of in‑group deviant members, the favorable evaluations of in‑group normative members, and support for the social control strategies would be significant predictors of social identification, but we were not sure about the relative predictive power of each independent variable in each experimental condition. We thus ran multiple regression analyses within each condition defined by Inter‑group Setting and Target’s Group. Post‑identification was the dependent measure, and pre‑identification, evaluative judgments toward each target, and support for the persuasive and punitive strategies were predictors1. Due to the exploratory nature of this analysis, we ran forward selection stepwise regressions (cf. Ho, 2006).

In the conditions in which participants evaluated out‑group members, the predictors did not explain post‑identification (criterion: probability of‑F‑to‑enter ≤ .05). In the threatening setting and evaluation of in‑group targets’ condition, post‑identification was predicted by support for the persuasive strategy, β = .60, t = 2.57, p = .025. In this condition, the more group members supported the persuasive civic group to deal with in‑group deviance, the more they identified after with the in‑group. In the secure setting and evaluation of in‑group targets’ condition, post‑identification was predicted by pre‑identification, β = .70, t = 4.95, p = .001, followed by the evaluation of the normative in‑group member, β = .36, t = 2.53, p = .032. The more favorably participants evaluated the normative in‑group member, the more they identified after with the in‑group. In sum, these results are consistent with our predictions. They show that evaluation of in‑group members and support for social control strategies allow individuals to warrant the positive value of the in‑group.

Discussion

The present results indicate that participants favored a normative in‑group member and derogated a deviant in‑group member in comparison to similar out‑group targets when the in‑group’s positive distinctiveness was threatened. In contrast, when the in‑group’s positive

1. In contrast to previous research on SGDT, we analyzed social control over in‑group deviance in general and not in particular group deviant members. Our present analysis is thus exploratory and does not focus on means. The complete data is available from the first author upon request.


distinctiveness was secure, the deviant in‑group member was less negatively evaluated than the deviant out‑group target. The non‑emergence of the black sheep effect in the latter condition seems at odds with Biernat, Vescio and Billings’ (1999) idea that derogation of deviant in‑group members ensues from the violation of higher positive expectancies about the in‑group than the out‑group.

Indeed, in light of this latter explanation, participants would derogate the in‑group deviant more than the out‑group deviant members, especially in the secure setting. Following the information that there are less instances of in‑group deviance than out‑group deviance, expectations should be more positive toward the in‑group than the out‑group, and a particular instance of in‑group deviance would be perceived as more unexpected than an instance of out‑group deviance. According to the expectancy violation explanation, this setting would thus be more fitting to a black sheep effect, than the threatening setting, where the in‑group deviant member is more in accordance to group‑based expectations than the out‑group deviant member.

In contrast, our results do not support the expected relationships between the evaluation of group targets, social control over in‑group deviance, and social identification. It is important to notice that previous research on subjective group dynamics measured the relationships between evaluations and social control toward particular group deviant members (Marques, Abrams, & Serodio, 2001; Pinto, et al., 2010). However, our study analyzed evaluations and social control toward different targets (in‑group or out‑group members, and in‑group deviance as a whole, respectively). This may have made the resulting patterns complex and non‑linear. Even so, two significant relationships emerged, which were in accordance to our predictions. In the secure setting, the more participants favored the in‑group normative member, the more they valued the in‑group afterwards. Also, in the threatening setting, the more participants advocated the persuasive strategy to deal with in‑group deviance, the more they assigned the in‑group a positive value afterwards. Perhaps these exploratory results justify follow‑up studies on the social motivations behind the punitive and socializing strategies over in‑group deviants. In addition, follow‑up studies should include measures about the estimations of in‑group deviant behavior and its impact on the in‑group’s image, which are important manipulation checks missing from our study.

Going back to our examples at the outset of this work, it is likely that the threats posed by those historical characters to their nations’ identities are relevant components for a more general explanation of the harshness of the social reactions they evoked. After all, Guy Fawkes was tortured and sentenced to death, amidst a vibrant civil conflict for the religious identity of England. Marshall Pétain was sentenced to death as part of France’s struggle to exorcise its collaborationist past and to restore a heroic identity. Mordechai Vanunu was sentenced to a 18 years imprisonment in a country whose national identity seems to be permanently under threat. Thus, these narratives seem to illustrate SGDT’s postulate that when social identity is threatened, group members can be strongly depreciative and hostile toward in‑group deviants.

References

Biernat, M., Vescio, T.K., & Billings, L.S. (1999). Black sheep and expectancy violation: Inte‑grating two models of social judgment. European Journal of Social Psychology, 29, 523‑542.

Branscombe, N.R., Wann, D.L., Noel, J.G., & Coleman, J. (1993). In‑group or out‑group extremity: Importance of the threatened social identity. Personality and Social Psychology Bulletin, 19, 381‑388, doi: 10.1177/0146167293194003.


Breakwell, G.M., Vignoles, V.L., & Robertson, T. (2003). Stereotypes and crossed‑category evaluations: The case of gender and science education. British Journal of Psychology, 94, 437‑455.

Cialdini, R.B., Reno, R.R., & Kallgren, C.A. (1990). A focus theory of normative conduct: Recycling the concept of norms to reduce littering in public places. Journal of Personality and Social Psychology, 58, 1015‑1026.

Eidelman, S., & Biernat, M. (2003). Derogating black sheep: Individual or group protection? Journal of Experimental Social Psychology, 39, 602, doi: 10.1016/s0022‑1031(03)00042‑8.

Haslam, S.A., & Turner, J.C. (1992). Context‑dependent variation in social stereotyping 2: The relationship between frame of reference, self‑categorization and accentuation. European Journal of Social Psychology, 22, 251‑277.

Hichy, Z., Mari, S., & Capozza, D. (2008). Pronorm and antinorm deviants: A test of the subjective group dynamics model. Journal of Social Psychology, 148, 641‑644.

Ho, R. (2006). Handbook of Univariate and Multivariate Data Analysis and Interpretation with SPSS. Boca Raton, Fl: Taylor & Francis Group.

Iyer, A., Schmader, T., & Lickel, B. (2007). Why individuals protest the perceived transgressions of their country: The role of anger, shame, and guilt. Personality & Social Psychology Bulletin, 33, 572‑587, doi: 10.1177/0146167206297402.

Jetten, J., Summerville, N., Hornsey, M.J., & Mewse, A.J. (2005). When differences matter: Inter‑group distinctiveness and the evaluation of impostors. European Journal of Social Psychology, 35, 609‑620, doi: 10.1002/ejsp.282.

Khan, S., & Lambert, A.J. (1998). Ingroup favoritism versus black sheep effects in observations of informal conversations. Basic & Applied Social Psychology, 20, 263‑269.

Levine, J.M., & Moreland, R.L. (1994). Group socialization: Theory and research. In W. Stroebe & M. Hewstone (eds.), European Review of Social Psychology (vol. 5). New York: Wiley.

Lewis, A.C., & Sherman, S.J. (2010). Perceived entitativity and the black‑sheep effect: When will we denigrate negative ingroup members? Journal of Social Psychology, 150, 211‑225.

Marques, J.M. (2010). Black sheep effect. In J.M. Levine & M. Hogg (eds.), Encyclopedia of Group Processes and Inter‑group Relations. Thousand Oaks, CA: Sage.

Marques, J., & Paez, D. (1994). The „black sheep effect”: Social categorization, rejection of ingroup deviates, and perception of group variability. European Review of Social Psychology, 5, 37‑68.

Marques, J.M., & Paez, D. (2008). Dynamique de groupes subjective: Un cadre théorique pour l’effet brebis galeuse. In R.‑V. Joule & P. Huguet (eds.), Bilans et perspectives en psychologie sociale (pp. 71‑115). Grenoble: Presses Universitaires de Grenoble.

Marques, J.M., & Yzerbyt, V.Y. (1988). The black sheep effect – Judgmental extremity towards ingroup members in inter‑group and intra‑group situations. European Journal of Social Psychology, 18, 287‑292.

Marques, J.M., Abrams, D., & Serodio, R.G. (2001). Being better by being right: Subjective group dynamics and derogation of in‑group deviants when generic norms are undermined. Journal of Personality and Social Psychology, 81, 436‑447.

Marques, J.M., Paez, D., & Abrams, D. (1998). Social identity and intragroup differentiation as subjective social control. In S. Worchel, J.F. Morales, D. Paez & J.‑C. Deschamps (eds.), Social Identity: International Perspectives (pp. 124‑141). Thousand Oaks, CA: Sage Publications, Inc.

Marques, J.M., Yzerbyt, V.Y., & Leyens, J.‑P. (1988). The “black sheep effect”: Extremity of judgments towards ingroup members as a function of group identification. European Journal of Social Psychology, 18, 1‑16.

Marques, J.M., Abrams, D., Paez, D., & Hogg, M. (2001). Social categorisation, social identi‑fication, and rejection of deviant group members. In Blackwell (ed.), Blackwell handbook of Social Psychology: Group Processes (pp. 400‑424). Oxford, UK: Blackwell.

Nesdale, D., & Brown, K. (2004). Children’s attitudes towards an atypical member of an ethnic in‑group. International Journal of Behavioral Development, 28, 328‑335.


Oakes, P.J., Haslam, S.A., & Turner, J.C. (1994). Stereotyping and Social Reality. Malden: Blackwell Publishing.

Pinto, I.R., Marques, J.M., Levine, J.M., & Abrams, D. (2010). Membership status and subjective group dynamics: Who triggers the black sheep effect? Journal of Personality and Social Psychology, 99, 107‑119, doi: 10.1037/a0018187.

Reno, R.R., Cialdini, R.B., & Kallgren, C.A. (1993). The transsituational influence of social norms. Journal of Personality and Social Psychology, 64, 104‑112.

Tajfel, H. (1978). Differentiation Between Social Groups: Studies in the Social Psychology of Inter‑group Relations. Oxford, UK: Academic Press.

Tajfel, H., & Turner, J. (1986). The social identity theory of inter‑group behavior. In S. Worchel & W.G. Austin (eds.), Psychology of Inter‑group Relations (pp. 7‑24). Chicago, Il: Nelson‑Hall Publishers.

Tajfel, H., Billig, M.G., Bundy, R.P., & Flament, C. (1971). Social categorization and inter‑group behaviour. European Journal of Social Psychology, 1, 149‑178.

Turner, J.C., Hogg, M.A., Oakes, P.J., Reicher, S.D., & Wetherell, M.S. (1987). Rediscovering the Social Group: A Self‑Categorization Theory. Cambridge, MA: Basil Blackwell.

Whitehead, G.I., Smith, S.H., & Eichhorn, J.A. (1982). The effect of subject’s race and other’s race on judgments of causality for success and failure. Journal of Personality, 50, 193‑202.

Documents

the effects of social identity threat