Previous research shows that men overall, in contrast to women, do not show a typical implicit in-group preference. One proposed explanation is greater interest in sex among males. If so, then gay males should show an implicit preference for males whereas straight males should prefer females. We tested this hypothesis using a modified version of the Brief Implicit Association Test on 38 gay and 65 straight males. The hypothesis was supported. As the majority of participants in previous studies on implicit gender attitudes are expected to be straight, this could contribute to the low implicit in-group bias among males.

This study confronted the classical idea that generalized prejudice is rooted in a cognitive tendency to sort reality into rigid and simple categories with the more recent idea that prejudice is shaped by moral intuitions. In a diverse Swedish sample (N = 430), moral absolutism was more strongly associated with generalized prejudice against derogated and dissident (but not dangerous) groups than were other aspects of intolerance of ambiguity. But there was little direct association between any aspect of intolerance of ambiguity and generalized prejudice once indirect relations through binding moral intuitions (which elevated prejudice) and individualizing moral intuitions (which decreased prejudice) had been taken into account. These findings suggest that intolerance of ambiguity is associated with generalized prejudice mainly insofar as it leads to a distinctly moral dichotomization of persons into categories such as insiders and outsiders, law-abiding citizens and deviants, and the righteous and the impure.

Tests like the Implicit Association Test (IAT), the Sorting Paired Features task (SPF) and the Brief IAT (BIAT), all to some degree contrast positively and negatively valenced categories against each other, in the task itself and/or in the scoring procedure. This makes it difficult to truly disentangle positive and negative associations, something that is important when measuring attitudes. A different approach was explored in a series of studies using a method similar to SPF and BIAT, but using a neutral category as a contrast against positive and negative associations. The method was tested using gender SPFs and gender BIATs, as well as various ethnicity SPFs. The results show support for the validity of the positive-neutral tests, but the results for the negative-neutral tests are less clear, and questions remain concerning reliability when using the latter test. Results on positive-netrual tests, on the other hand, showed some support for its validity, for instance was scores on a gender BIAT related to the level of appreciation of sexist jokes. Further studies using large samples are needed to determine the psychometric of this method.

This thesis explores in-group bias control. It is well-known that people tend to have extra liking for people they identify themselves with. An extra positive in-group attitude may cause discrimination, even in the absence of any negative attitudes towards the out-group. At the same time are many people motivated to behave and present themselves as unbiased. We study whether people attempt to control favourable in-group evaluations to a more modest level due to concerns about being and/or appearing biased.Empirical studies are presented in three papers. Paper 1 studied how the presence of either a member of the in-group or a member of the out-group influenced how favourably the in-group was evaluated. In a between-groups design, participants tended to rate their in-group more favourably when they gave their evaluations verbally to an in-group member rather than to an out-group member. In paper 2, a scale intended to measure individual differences in motivation to control in-group bias was constructed, and its psychometric properties were tested. Inspired by the literature on out-group bias correction, two subscales where items either referred to internal motivation to control in-group bias or to external motivation to control in-group bias were created. Exploratory and confirmatory factor analyses gave some support to the notion of two distinct, but related factors. However, the temporal reliability of the scales was lower than expected. The scales showed some relations to other theoretically relevant measures, but more reliable scales are needed in order to further test their validity.In paper 3 corrections in evaluations of people of different ethnicity and sex was explored. Participants evaluated pictures of Black males, Black females, White males, White females, Middle Eastern males and Middle Eastern females. When participants were given a reminder of the risk of bias, a three-way interaction effect was evident across three experiments. The correction patterns varied somewhat between the experiments, but an overall pattern was that a cue to the risk of bias led to more favourable ratings of some groups (Black men, Middle Eastern men and White women) and a less favourable rating of other (White men). A final study explored a normative account of the findings from the experiments, and a clear pattern emerged where people considered it more important to avoid overrating White men compared to the other groups, whereas it was considered less important to avoid underrating white men compared to the other groups under study. Implications of the results and some thoughts regarding future research are discussed. (Less)

A measure of individual differences in the motivation to control in-group bias ( favouritism) was created. As in Plant and Devine’s (1998) measure of out-group bias, one subscale referred to an internal motivational source and the other to an external motivational source. The psychometric properties of the measure were tested across four samples. The results indicate that the measure reliably captures individual differences in the motivation to control in-group bias, and that a distinction between internal and external motivational sources of control can be made. However, the convergent and discriminant validity of both subscalesneeds to be examined further before the measure can be considered launchable. Implications for research on intergroup bias are discussed.

Correction of judgments of people of different ethnicities and sexes was exploredby exposing participants to cues to the risk of making biased judgments (from explicitwarnings to subtle hints). In three experiments, a three-way interaction was revealed, wherethe effect of a cue to bias varied as a function of both the ethnicity and sex of the targetperson. Some targets (White males) were generally rated less favorably when judges werereminded of bias, whereas other targets (Black males, Middle Eastern males and Whitefemales) were generally rated more favorably, indicating bidirectional correction. Finally, anormative account of the results was explored. In a pattern consistent with the experimentalresults, it was considered more important to avoid overrating White men than all othergroups, and more important to avoid underrating all other groups than White men. Theresults are discussed in relation to theories of correction and intergroup bias.

Much of what we know concerning impression formation is based on experimental methods where the participant receives a list of traits or behaviors and is asked to make trait judgments or meta-cognitive judgments. The present study aimed to put some well-known effects from the impression formation literature to a test in a more dynamic computerized environment, more akin to many real world impression formation scenarios. In three studies participants were introduced to multiple target persons. They were given information about the target persons' behavior, one at a time, while making ratings of their warmth and competence, and their probability of performing related behaviors in the future. In neither of the studies the negativity effect of warmth or the positivity effect of competence were reproduced.