IN THE HEAT OF THE MOMENT: AN EXPERIMENTAL INVESTIGATION OF INTERGROUP DISCRIMINATION AS A MOTIVATED RESPONSE TO REPAIR A TRANSIENT NEGATIVE MOOD

Using a variant of the minimal group paradigm, five experiments examined the effects of transient mood, specifically negative mood, on intergroup discrimination. Despite recent interest in the role of affect in social cognition (e.g., Baron, 1987; Erber & Erber, 2001; Forgas & Bower, 1987; Giner-Sorolla, Mackie & Smith, 2007), there has been little systematic investigation of the role of transient mood in intergroup processes. As an exception, Forgas and Fiedler (1996) reported evidence that the combination of negative mood and a personally involving group membership motivated discriminatory behaviour on the Tajfel (1970) matrices. The central aim of the thesis was to extend the research findings of Forgas and Fiedler via investigation of the joint effects of group status and transient negative mood on discrimination with a focus on the mechanism believed to underpin bias for negative mood respondents: motivated mood repair. Experiment One used a 2 x 3 experimental design. Participants (N = 93) were categorised into groups that varied in terms of mood (negative, neutral, positive) and relative group status (high, low) before distributing points to ingroup and outgroup members using the Tajfel (1970) matrices. Results indicated that, overall, high status group members were more discriminatory than members of low status groups and members in a negative mood were more discriminatory than members in an induced positive or negative mood. This suggested, as per Forgas and Fiedler (1996), that transient negative mood is an important factor in intergroup discrimination, at least when the ingroup imbues positive self regard. Most discrimination was evidenced for high status participants in a negative mood. Results examining between-groups differences in perceived mood repair and task enjoyment indicated that respondents in a negative mood high status condition, those who discriminated most, also reported enhanced mood improvement and task enjoyment. Further, mediational analyses revealed that, among members of high status groups, the effect of mood on perceived mood repair and task ease was no longer significant when ingroup bias was included as a covariate. Taken together, results from Experiment One provided initial evidence for motivated processing in the service of mood repair. Experiment Two used a one-way experimental design (N = 87) to examine more closely the potential for intergroup discrimination to effect mood repair. Respondents in a negative mood-high status condition completed the Tajfel (1970) matrices either after standard instructions for matrix completion or after instructions explicitly indicating that they should use a fairness, an outgroup bias, or an ingroup bias strategy. Results provided further evidence of a tendency to discriminate in favour of the ingroup in negative mood high status contexts. Results also confirmed that, relative to other possible responses (e.g., outgroup bias), the display of ingroup bias promoted mood repair, apparently via highly targeted, motivated processing rather than via slower, deliberative processing (cf. Forgas & Fiedler, 1996). The third experiment in the series sought to explicitly test whether motivated mood repair acted as the driver for accentuated ingroup bias or whether mood repair was simply a collateral benefit of discrimination. Participants (N = 77), all of whom were classified as high status group members, were induced into either a negative or neutral mood state. Then, borrowing Manucia, Baumann and Cialdinis (1984) mood freeze pill paradigm, respondents were told that, having ingested a pill, their current mood could or could not be altered. Respondents in the fixed mood condition did not engage in the level of bias typically expected from sad participants and, further, they did not enjoy the same level of mood repair as labile mood respondents who showed significantly greater ingroup bias. Thus mood repair was identified as the direct driver of discrimination in this context. The fourth experiment (N = 92) examined the role of allocation norms (ingroup favouritism, outgroup favouritism) and mood (negative, positive) on discrimination and mood repair among members of high status groups. As predicted negative mood participants displayed more discrimination than their positive mood counterparts and participants advised of a norm supportive of ingroup favouritism displayed accentuated discrimination compared with their peers advised of a norm for outgroup favouritism. A complex but complementary pattern was reported on the mood repair and task ease measures, indicating that those who were supported by a norm for discrimination found the task easier and displayed more mood repair. The fifth and final experiment (N = 80) sought to examine the role of two simple behavioural checks on manifested discrimination for negative mood, high status respondents. A 2 (mood awareness: not aware, aware) X 2 (response latency: spontaneous, delayed) experimental design was employed and results confirmed the efficacy of both manipulations in attenuating bias on the minimal group matrices. Presumably, mood awareness and longer decision timeframes triggered less urgent mood calibration efforts (e.g., Forgas, Ciarrochi, & Moylan, 2000). The final chapter presents key theoretical and practical implications of the research findings, including: confirmatory support for the notion that, in a negative mood high status context, mood repair is a potent driving force for discriminatory behaviour; specific evidence that, in selecting a cognitive processing strategy, the perceiver is sensitive to a hierarchical series of cueing variables (including status, relevance and mood); initial indications that relatively simple interventions can attenuate the typical effects of negative mood on discrimination; and the continued importance of social norms and education that emphasise the potential influence of mood on intergroup behaviour and the social value of parity in intergroup relations. Finally, strengths and limitations of the experimental approach are acknowledged and suggestions proffered for future applied testing