CURRENT RESEARCH IN SOCIAL PSYCHOLOGY

Social identity theory (SIT) has dominated much of the work carried on inter-group relations over the past two decades. An implicit assumption of the theory is that category members engage in various forms of inter-group discrimination in order to achieve and maintain self-esteem (Abrams & Hogg, 1988; Tajfel & Turner, 1979). The research carried out to investigate this and the related assumption, that low or threatened self-esteem can enhance inter-group discrimination, reveals little unambiguous support for either premise (see Abrams & Hogg, 1988; Hunter, Platow, Bell, Kypri & Lewis, 1997, Rubin & Hewstone, 1998, for reviews). In light of the complex and often contradictory nature of the findings discerned in this field a growing number of influential researchers have now begun to question the pivotal motivational role assigned to self-esteem within the SIT framework (Brown, 1995; Hogg & Abrams, 1993).

DISCUSSION

In addressing this issue, it is important to acknowledge a number of criticisms that have been leveled at the research conducted in this area. One criticism with particular relevance to the current study relates to the fact that many researchers, when attempting to assess how inter-group discrimination affects self-evaluation, have utilized global measures of self-esteem. The use of such instruments to examine predictions derived from SIT is highly problematic. According to SIT the self-concept is multidimensional. Depending on how the self is experienced, components of the self (or self-descriptions) may be related to either personal or social identities (Turner, Oakes, Haslam & McGarty, 1994). Components of the self experienced at the level of the individual relate to personal identities. Components of the self experienced at the level of the group relate to social identities. Stressing the social aspects of identity, SIT posits that, in the relevant context, those domains of the self associated with social identities will become more salient than those domains of the self associated with personal identities. As a result, therefore, when any given social identity becomes salient (e.g. Baptist) people will tend to define and evaluate themselves, not in terms of their overall personal identity but, in terms of those components of the self (e.g. religious, honest) related to their social identities. One consequence of this is that global measures of self-esteem, which are designed to provide a generic measure of personal self-worth cannot accurately assess those aspects of the self associated with social identity.

As a result of this reasoning, researchers have now begun to seek alternative methods by which to more accurately examine social identity based self-esteem (see Rubin & Hewstone, 1998 for a review). One possibility has been outlined by Hunter and his colleagues (Hunter, Platow, Howard & Stringer, 1996, Hunter et al., 1997). Combining Marsh’s work on the multidimensional self-concept (e.g. Marsh, 1992, 1993) with social identity and self-categorization theory (e.g. Turner et al., 1994) the aim of this approach is to examine the self descriptions (e.g. sportsmanlike, happy-go-lucky) which may be subsumed under particular social identities (i.e. Australian). Recent research would tend to support this perspective. Thus, a number of studies have shown that, when the members of meaningful social categories display group based bias, it is, not global but, domain specific self-esteem that is affected. For example, in two studies utilizing gender categories, Hunter et al. (1997) found that biased evaluations of in-group and out-group targets led to changes in verbal and physical self-esteem but not global self-esteem. Identical findings were reported in a sample comprising Northern Irish participants. In this study, Hunter et al. (1996) again assessed global and domain specific self-esteem prior to and following the manifestation of evaluative in-group bias. No effects were found for global self-esteem. The esteem in which both Protestants and Catholics held specific self-images (e.g. physical appearance, religiosity, honesty, verbal and academic ability) was, however, found to increase after they engaged in evaluative in-group bias.

The research carried out by Hunter et al., demonstrates one way in which our understanding of the link between self-esteem and inter-group discrimination may be advanced. Such work does, however, raise a number of other issues. Specifically, since there are an infinite number of possible self-esteem domains (Hogg & Abrams, 1988; Marsh, 1993) which may plausibly be subsumed under any specific social identity, it is important for both theoretical and practical reasons (e.g. Abrams, 1996; Brown, 1995) that we are able to identify those particular components of the self which are likely to be related to inter-group differentiation. The particular self-esteem domains likely to increase following the display of in-group bias was not apparent in the studies carried out by Hunter et al. (1996, 1997). One possible explanation for this state of affairs may, however, be derived from recent developments in social identity and self-categorization theory (Hogg & Abrams, 1988; Turner, Hogg, Oakes, Reicher & Wetherell, 1987). According to this framework the functioning of the self-concept is context dependent. Particular components of the self are activated as a logical function of the "interaction between the characteristics of the perceiver and the situation" (Turner et al., 1987 p. 44). As a consequence, there is both continuity and variability in the contents of self-perception. When there is variability in the inter-group situation the attributes associated with social category membership change. When there is stability in the inter-group situation the attributes associated with social category membership remain stable (see Oakes, Haslam & Turner, 1994; Turner et al., 1994 for reviews).

Design
Participants assigned to the experimental condition were given the opportunity to evaluate in-group (Christian) and out-group (Atheist) target actors who behaved either positively or negatively. This formed a 2 (group membership of target actor: in-group/out-group) x 2 (behavior outcome: positive or negative) repeated measures design. Participants assigned to the first control condition completed the same evaluative tasks as those in the experimental condition with the exception that target actors were identified as belonging to one of two out-groups (i.e. Sephardic or Hasidic Jews). The dependent variables (administered after the inter-group evaluative tasks) were scales assessing global, religious and mathematical self-esteem. In order to provide an assessment of pre-test levels of self-esteem, participants assigned to the second, baseline, control condition completed the measures of self-esteem without previously taking the inter-group evaluative tasks.

Materials and Procedure
The study was introduced as being concerned with self-perception, social judgement and behavior. Participants were told that during the course of the investigation they would complete a number of response booklets and then engage in a short behavioral exercise. In an attempt to facilitate social identity salience (and thereby preclude this variable as a potential cause of self-esteem change, see Abrams & Hogg, 1988 for a review) participants were informed that the study was specifically concerned with groups of Christians and Atheists. To further heighten this effect, and also control anticipated interaction time amongst in-group and out-group members, attention was drawn to the behavioral exercise that was to be carried out at the end of the study. This (bogus) exercise was said to involve a 5-minute interaction with Christians (in-group members) and a 5-minute interaction with Atheists (out-group members). Atheists were said to be involved in an identical experiment being carried out simultaneously in an adjacent room. To ensure anonymity of responding, participants chose a code number from a box that was passed round the room. Participants were required to record this code number and the social group (i.e. Christian) to which they belonged on each of their response booklets. Communication amongst participants was discouraged whilst the study was in progress.

Booklet 1: Evaluative bias and distractor tasks
Participants assigned to the experimental condition were presented with a number of vignettes. Each vignette was one paragraph long and depicted a situation in which the participant was asked to imagine that an in-group or an out-group target actor had behaved either positively or negatively towards them. Target actors were identified through name and specific reference to social category membership (e.g. ‘Greg a Christian’ or ‘Simon an Atheist’).1 The vignettes, 12 in total, comprised one example each of a target actor who either: (a ) helped or ignored the participant, after he or she had been involved in a minor accident, (b) was generous to or cheated the participant and (c) helped or refused help to the participant when he or she had been caught in the rain. The vignettes were presented in a single random order. Following the presentation of each vignette, participants were then asked to evaluate the target actor in question (e.g. How would you evaluate Greg/Simon?). Evaluations were assessed using seven-point Likert scales (7-very positive to 1-positive, see Dovidio & Fazio, 1992 for a discussion of the strengths of this approach). Participants assigned to the first control condition completed the same evaluative tasks as those in the experimental condition with the exception that target actors were identified as belonging to one of two out-groups -- Sephardic or Hasidic Jews. Participants assigned to the second, baseline, control condition did not complete the evaluative tasks. Twenty minutes were given to complete these tasks.

Booklet 2: Self-esteem measures
Immediately following the completion of the first response booklet, a second booklet was administered. This contained the general self-evaluation, religious and mathematical self-esteem sub-scales of the Self-Description Questionnaire III (SDQ III).2 In its entirety, the SDQ III is designed to measure 13 separate domains of self-esteem (Marsh, 1992) and is one of the most extensively validated self-esteem instruments currently available (Byrne, 1996, p. 204). Each domain of self-esteem assessed by the SDQ III has high internal consistencies (median alpha's of between .80 and .90) and test-retest reliabilities (median r's of .87). Correlations between domains are low (median r's of .10). These findings mean that sub-scales measuring each self-esteem domain may be used either separately or in combination (Marsh, 1992). Each self-esteem domain is measured by a single scale consisting of 10 or 12 items. The general self-evaluation sub-scale is derived from the Rosenberg (1965) global self-esteem scale. It is therefore similar to those used in other studies which have sought to examine the link between various forms of inter-group discrimination and self-esteem (e.g. Hunter et al., 1997). An example of the content of the three sub-scales used in the current investigation are as follows: "Overall, I am pretty accepting of myself (general), "I am a better person as a result of my religious/spiritual beliefs" (religion) and "I am quite good at mathematics" (mathematical ability). All answers are recorded on an 8 point Likert scale (1- Definitely False, 8-Definitely True). Higher scores reflect more positive levels of self-esteem. Half of the items in each scale are scored in the reverse order. Participants are required to respond to all questions on the basis of how they "now feel" and "not as [they] usually feel."

Manipulation checks
In order to examine potential differences in social identity salience across experimental conditions (M=23.50), the first control condition (M=23.53) and the baseline control conditions (M=24.02) a one way analysis of variance (ANOVA) was conducted. No significant effects were discerned (F(2, 138)=.25, p>.78). This indicates that social identity was similarly salient amongst experimental and control participants. To ensure participants differed in how they evaluated religion and mathematics to the Christian in-group a 2 x (condition: experimental/control) x 2 (dimension: religious/mathematical) mixed model ANOVA was conducted. The last factor was repeated. A main effect was found for dimension. As expected, this analysis revealed that, for the participants in the present sample, religion was judged to be more important to the Christian in-group than mathematical ability (M=6.57 v M=2.39, F(2, 138)=1060.44, p<.001). There were no other significant effects.

Evaluative in-group bias
Evaluations of in-group and out-group target actors were analyzed by means of a 2 (group membership of target actor: in-group/out-group) x 2 (behavior outcome: positive or negative) repeated measures analysis of variance (ANOVA). After the standard methodology in this area (see Islam & Hewstone, 1993) responses were collapsed so that one overall score was computed for each episode of positive and negative in-group and out-group behavior. Cell means can be seen in Table 1. Higher scores indicate more positive evaluations. A main effect was found for behavior outcome (F(1, 47)=98.28, p<.001). Positive behaviors were evaluated more positively than negative behaviors (M=5.85 v M=4.18). A further main effect was found for target group of evaluation (F(1, 47)=26.00, p<.001). In-group members were evaluated more positively than out-group members (M=5.38 v M=4.65). Planned comparisons revealed that (a) in-group targets who behaved positively were evaluated more highly than out-group targets who behaved positively (t(47)=5.40, p<.0005) and (b) in-group targets who behaved negatively were evaluated more positively than out-group targets who behaved negatively (t(47)=3.16, p<.005). Both these effects were significant using Dunn’s test (Bonferroni t, critical alpha value 2.97, p<.01). There were no other significant effects.

Table 1.  Mean Evaluative Ratings for the Positive and Negative Behavior of In-group and Out-group Targets

Self-esteem
To examine differences in general, religious and mathematical self-esteem among those assigned to experimental and control conditions, separate one way analyses of variance (ANOVAs) were conducted. Cell means can be seen in Table 2. Contrary to expectations no significant effects were found for religious self-esteem (F(2, 138)=1.23, p>.29). An effect was, however, found for global self-esteem (F(2, 138)=3.79, p<.03). To investigate this effect further a series of post hoc comparisons were conducted. This analysis revealed that participants in the experimental condition had lower levels of global self-esteem than those in the first control (t(91)=2.31, p<.03) and those in the baseline control conditions (t(94)=2.29, p<.03). Both these effects were significant using Dunn’s correction (critical alpha value 2.29, p<.05). An effect approaching significance was found for mathematical self-esteem (F(2, 138)=2.94, p<.06). Post hoc comparisons were conducted to assess this effect further. This analysis revealed participants in the experimental condition to have lower levels of mathematical self-esteem than those in the baseline control condition (t(94)=2.34, p<.05). This effect was also significant using Dunn’s correction (critical alpha value 2.29). There were no other significant effects.

Table 2.  Experimental and Control Participants Global and Domain Specific Self-Esteem

General Discussion
One hypothesis was tested in this experiment. This predicted that category members (i.e. Christians) would experience an increase in that domain of self-esteem judged to be more important to the in-group (religion), following the display of evaluative in-group bias. No support was found for this hypothesis. In the experimental condition, regardless of whether they behaved positively or negatively, participants evaluated in-group targets (i.e. Christians) more highly than out-group targets (i.e. Atheists). Following the manifestation of these biases participants failed to experience an increase in that domain of self-esteem judged to be more important to the in-group (i.e. religion). Contrary to expectations those Christians who displayed evaluative in-group bias showed lower levels of global and mathematical self-esteem (although, with respect to the latter domain this was only in comparison to the baseline control condition).

These findings are at odds with those reported by Hunter and his colleagues (1996, 1997). Hunter et al., found that the display of evaluative in-group bias lead to increases in domain specific self-esteem. In this sense it may be argued that the failure of the current investigation to replicate the work of Hunter et al, may be a joint consequence of (a) the relatively low worth of mathematical self-esteem to participants and (b) that (given the Christian emphasis on individual responsibility) the religious domain of self-esteem (as assessed by the SDQ III) may be chronically accessible at the personal rather than the social level of identity (see also Ng & Wilson, 1989). Similarly, it should also be acknowledged that we cannot, of course, discount the possibility that participants may have experienced elevated feelings of self-worth on some other domain of self-esteem (e.g. honesty, academic ability) which was not included in the present study. Lowered levels of self-esteem following the clear expression of in-group bias are, however, particularly difficult to reconcile with the SIT assumption that category members engage in various forms of inter-group discrimination in order to enhance their self-esteem.

One possible explanation for these findings can of course be easily offered through reference to the social norms of the Christian participants. An important part of the Christian doctrine entreats believers to love others ‘as thy self’. Consequently, by showing bias in favor of their own group it is possible that category members may have experienced feelings of shame and thus subsequently lowered self-esteem. While such an interpretation is plausible with respect to the present study, it is important to remember that the findings discerned in the current investigation are not the first to record decreased self-esteem amongst those who shown in-group bias. Indeed, similar findings have been reported by a number of other researchers (e.g. Hogg, Turner, Nascimento-Schulze, & Spriggs, 1986, experiment 2; Vickers, Abrams & Hogg, 1988). Likewise, in a recent study, Branscombe & Wann (1994) found lowered levels of collective self-esteem amongst U.S. college students who derogated out-groups comprising French, South African and Chinese nationals.

1.  To control for any potential bias in the naming of stimulus persons a separate pilot study was conducted (see Kasof, 1993 for a review). Participants (N=82) evaluated the pleasantness of the Christian and Atheist names that were to be used in the current investigation (e.g. Greg, Simon). Responses were scored on a 7-point Likert scale (7-pleasant, 1-unpleasant). A repeated measures t-test revealed that there were no significant differences (M=4.19 v M=4.27, t(81)=.76, p>.45).

2.  The self-esteem domain of religion was included on the basis of the assumption that it would be judged to be relatively important to the Christian in-group. The domain of mathematical self-esteem was included as a result of pilot testing. To determine which self-esteem domains were likely to be deemed as being important and unimportant in the current context a separate pilot test was carried out. Respondents (N=130), recruited from the general student population, were asked to rate the importance of 12 SDQ III self-esteem domains in the current context (physical appearance, physical ability, opposite sex relations, same sex relations, parental relations, emotional stability, spirituality verbal ability, mathematical ability, academic ability creative ability and honesty). Ratings were assessed on 7 point Likert scales (1-Unimportant, 7-Important). Physical appearance (M=5.71) and Emotional stability (M=5.58) received the highest evaluations. Mathematical ability received the lowest rating (M=3.69). Consequently we decided to incorporate the latter into the present investigation.

Abrams, D. (1996). "Social identity, self as structure and self as process." In W. P. Robinson (Ed), Social Groups and Identities: Developing the Legacy of Henri Tajfel. Oxford: Butterworth/Heinemann.

Abrams, D. & Hogg, M. A. (1988). "Comments on the motivational status of self-esteem in social identity and intergroup discrimination." European Journal of Social Psychology, 18, 317-334.

Branscombe, N. R. & Wann, D. L. (1994). "Collective self-esteem consequences of out-group derogation when a valued social identity is on trial." European Journal of Social Psychology, 24, 641-657.

Brown, R. (1995). Prejudice: Its Social Psychology. Oxford: Blackwell.

Byrne, B. M. (1996). Measuring Self-Concept Across the Life Span: Issues and Instrumentation. American Psychological Association.

Dovidio, J. F. & Fazio, R. H. (1992). "New technologies for the direct and indirect assessment of attitudes." In J. M. Tanur (Ed.), Questions About Questions: Inquiries into the Cognitive Bases of Surveys. New York: Russell Sage Foundation.

Harter, S. (1986). "Processes underlying the construction, maintenance and enhancement of the self-concept in children." In J. Suls & A. G. Greenwald (Eds), Psychological Perspectives on the Self (vol 3). Hillsdale, NJ: Erlbaum.

Hogg, M. A. & Abrams, D. (1988). Social Identifications: A Social Psychology of Intergroup Relations and Group Processes. London: Routledge.

Hogg, M. A. & Abrams, D. (1993). "Towards a single-process uncertainty reduction model of social motivation in groups." In M. A. Hogg & D. Abrams (Eds), Group Motivation: Social Psychological Perspectives. London: Harvester/Wheatsheaf.

Hogg, M. A., Turner, J. C., Nascimento-Schulze, C., Spriggs, D. (1986). "Social categorization, intergroup behaviour and self-esteem: Two experiments." Revista de Psicologia Social, 1, 23-37.

Hunter, Platow, M. J., Bell, L. M., Kypri, K., Lewis, C. A. (1997). "Intergroup bias and self-evaluation: Domain specific self-esteem, threats to identity and dimensional importance." British Journal of Social Psychology, 36, 405-426.

Hunter, J. A., Platow, M. J., Howard, M. L. & Stringer, M. (1996). "Social identity and intergroup evaluative bias: Realistic categories and domain specific self-esteem in a conflict setting." European Journal of Social Psychology, 26, 631-647.

Islam, M. R. & Hewstone, M. (1993). "Intergroup attributions and affective consequences in majority and minority groups." Journal of Personality and Social Psychology, 64, 936-950.

Kasof, J. (1993). "Sex bias in the naming of the stimulus persons." Psychological Bulletin, 113, 140-163.

Leary, M. R. & Downs, D. L. (1995). "Interpersonal functions of the self-esteem motive: The self-esteem system as a sociometer." In M. J. Kernis (Ed), Efficacy, Agency and Self-Esteem. New York: Plenum Press.

Luhtanen, R. & Crocker, J. (1992). "A collective self-esteem scale: Self-evaluation of one’s social identity." Personality and Social Psychology Bulletin, 18, 302-318.

Marsh, H. W. (1992). Manual for the Self-Description Questionnaire III. University of Western Sydney: Australia.

Marsh, H. W. (1993). "Academic self-concept: Theory measurement and research". In J. Suls (Ed.), Psychological Perspectives on the Self (vol 4). Hillsdale, NJ: Erlbaum.

Ng, S. H. & Wilson, S. (1989). "Self-categorization theory and belief polarization among Christian believers and atheists." British Journal of Social Psychology, 28, 47-56.

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

Rosenberg, M. (1965). Society and the Adolescent Self-Image. Princeton, NJ: Princeton University Press.

Rubin, M. & Hewstone, M. (1998). "Social identity theory’s self-esteem hypothesis: A review and some suggestions for clarification." Personality and Social psychology Review, 2, 40-62.

Tajfel, H. & Turner, J. C. (1979). "An integrative theory of intergroup conflict." In W. G. Austin & S. Worchel (Eds), The Social Psychology of Intergroup Relations. Monterey: Brooks/Cole.

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

Turner, J. C., Oakes, P., Haslam, S. A. & McGarty, C. (1994). "Self and collective: Cognition and social context." Personality and Social Psychology Bulletin, 20, 454-463.

John A. Hunter completed his B.S. and Ph.D. at the University of Ulster in Northern Ireland.  Since completing, he has taken a lectureship at the University of Otago in New Zealand.  His main research interests are concerned with the motivational basis of inter-group discrimination.  He may be reached at the following address: Psychology Department, University of Otago, Dunedin, P.O. Box 56, New Zealand, Tel 64 3 479-7619. Fax 64 3 479-8335. E-mail: JHunter@rivendell.otago.ac.nz