Gender bias in the Trends in Mathematics and Science Study 2003 (TIMMS) for Canadian students /

This study is a secondary data analysis of the Trends in Mathematics and Science Study 2003 (TIMSS) to determine if there is a gender bias, unbalanced number of items suited to the cognitive skill of one gender, and to compare performance by location. Results of the Grade 8, math portion of the test were examined. Items were coded as verbal, spatial, verbal /spatial or neither and as conventional or unconventional. A Kruskal- Wallis was completed for each category, comparing performance of students from Ontario, Quebec, and Singapore. A Factor Analysis was completed to determine if there were item categories with similar characteristics. Gender differences favouring males were found in the verbal conventional category for Canadian students and in the spatial conventional category for students in Quebec. The greatest differences were by location, as students in Singapore outperformed students from Canada in all areas except for the spatial unconventional category. Finally, whether an item is conventional or unconventional is more important than whether the item is verbal or spatial. Results show the importance of fair assessment for the genders in both the classroom and on standardized tests.

Perceptions of intergroup bias : the roles of social projection and meta-stereotypes

Underlying intergroup perceptions include processes of social projection (perceiving personal traitslbeliefs in others, see Krueger 1998) and meta-stereotyping (thinking about other groups' perceptions of one's own group, see Vorauer et aI., 1998). Two studies were conducted to investigate social projection and meta-stereotypes in the domain of White-Black racial relations. Study 1, a correlational study, examined the social projection of prejudice and 'prejudiced' meta-stereotypes among Whites. Results revealed that (a) Whites socially projected their intergroup attitudes onto other Whites (and Blacks) [i.e., Whites higher in prejudice against Blacks believed a large percentage of Whites (Blacks) are prejudiced against Blacks (Whites), whereas Whites low in prejudice believed a smaller percentage of Whites (Blacks) are prejudiced]; (b) Whites held the meta:..stereotype that their group (Whites) is viewed by Blacks to be prejudiced; and (c) prejudiced meta-stereotypes may be formed through the social projection of intergroup attitudes (result of path-model tests). Further, several correlates of social projection and meta-stereotypes were identified, including the finding that feeling negatively stereotyped by an outgroup predicted outgroup avoidance through heightened intergroup anxiety. Study 2 replicated and extended these findings, investigating the social projection of ingroup favouritism and meta- and other-stereotypes about ingroup favouritism. These processes were examined experimentally using an anticipated intergroup contact paradigm. The goal was to understand the experimental conditions under which people would display the strongest social projection of intergroup attitudes, and when experimentally induced meta-stereotypes (vs. other-stereotypes; beliefs about the group 11 preferences of one's outgroup) would be most damaging to intergroup contact. White participants were randomly assigned to one of six conditions and received (alleged) feedback from a previously completed computer-based test. Depending on condition, this information suggested that: (a) the participant favoured Whites over Blacks; (b) previous White participants favoured Whites over Blacks; (c) the participant's Black partner favoured Blacks over Whites; (d) previous Black participants favoured Blacks over Whites; (e) the participant's Black partner viewed the participant to favour Whites over Blacks; or (£) Black participants previously participating viewed Whites to favour Whites over Blacks. In a defensive reaction, Whites exhibited enhanced social projection of personal intergroup attitudes onto their ingroup under experimental manipulations characterized by self-concept threat (i.e., when the computer revealed that the participant favoured the ingroup or was viewed to favour the ingroup). Manipulated meta- and otherstereotype information that introduced intergroup contact threat, on the other hand, each exerted a strong negative impact on intergroup contact expectations (e.g., anxiety). Personal meta-stereotype manipulations (i.e., when the participant was informed that her/ his partner thinks s/he favours the ingroup) exerted an especially negative impact on intergroup behaviour, evidenced by increased avoidance of the upcoming interracial interaction. In contrast, personal self-stereotype manipulations (i.e., computer revealed that one favoured the ingroup) ironically improved upcoming intergroup contact expectations and intentions, likely due to an attempt to reduce the discomfort of holding negative intergroup attitudes. Implications and directions for future research are considered.

Subjective Ambivalence and the Expression of Anti-Gay Bias

Univalent attitudes toward gay people have been widely studied, but no research to date has examined ambivalent (i.e., torn, conflicted) attitudes toward gay people. However, the Justification-Suppression Model (JSM; Crandall & Eshleman, 2003) proposes that ambivalence leads to biased expressions through intrapsychic processes which facilitate biased expression, particularly in contexts presenting strong justifications for expressing prejudice and weak pressures to suppress prejudice. I test these implications in the context of bias toward gay people. In Study 1, the measurement of ambivalence is examined in terms of both subjective ambivalence (i.e., the reported experience of “torn” attitudes) and calculated ambivalence (i.e., mathematical conflict between positive and negative attitude components). I find that higher subjective ambivalence is only associated with more negative attitudes toward gay people (and not positive attitudes toward gay people), and that higher subjective ambivalence predicts less gay rights support even after taking negative and positive attitudes toward gay people into account. Further, higher subjective ambivalence is associated with ideological opposition to gay people and more negative intergroup emotions (e.g., intergroup disgust). These findings suggest it is valuable to examine the unique component of subjective ambivalence separate from univalent negativity. Because calculated ambivalence measures are mathematically dependent upon a univalent negative measure, they cannot be examined separately from negativity. Therefore, subjective ambivalence is the focus of Study 2. The main goals of Study 2 were to determine why and when subjective ambivalence is related to bias. I examined the extent to which the negative relation between subjective ambivalence and opposition to anti-gay bullying can be accounted for by lower intergroup empathy and lower collective guilt, which may facilitate the expression of bias in keeping with the JSM. The relation between subjective ambivalence and anti-gay bullying opposition was examined within four social contexts based on a 2 (high vs. low offensiveness) x 2 (normatively unjustified vs. normatively justified) manipulation. I expected that higher subjective ambivalence would be most strongly related to lower intergroup empathy and collective guilt when there are the strongest justifications for bias expression, and that lower intergroup empathy and collective guilt would lead to less opposition to anti-gay bullying. Higher subjective ambivalence predicted less anti-gay bullying opposition. After accounting for positivity and negativity, the direct effect of subjective ambivalence was no longer significant, yet subjective ambivalence uniquely predicted intergroup empathy, which in turn predicted less anti-gay bullying opposition. These findings provide evidence that subjective ambivalence is largely negative in nature, but also presents evidence for a unique component of subjective ambivalence (separate from univalent attitudes) associated with low intergroup empathy and negativity. In contrast to previous research, I found very little evidence for the context-dependency of subjective ambivalence. Further research on subjective ambivalence, including subjective ambivalence toward other social groups, may expand our understanding of the factors leading to biased expressions.

Subliminal Exposure to Empathy Primes: A New Method to Reduce Intergroup Bias?

An abundant literature has demonstrated the benefits of empathy for intergroup relations (e.g., Batson, Chang, Orr, & Rowland, 2002). In addition, empathy has been identified as the mechanism by which various successful prejudice-reduction procedures impact attitudes and behaviour (e.g., Costello & Hodson, 2010). However, standard explicit techniques used in empathy-prejudice research have a number of potential limitations (e.g., resistance; McGregor, 1993). The present project explored an alternative technique, subliminally priming (i.e., outside of awareness) empathy-relevant terms (Study 1), or empathy itself (Study 2). Study 1 compared the effects of exposure to subliminal empathy-relevant primes (e.g., compassion) versus no priming and priming the opposite of empathy (e.g., indifference) on prejudice (i.e., negative attitudes), discrimination (i.e., resource allocation), and helping behaviour (i.e., willingness to empower, directly assist, or expect group change) towards immigrants. Relative to priming the opposite of empathy, participants exposed to primes of empathy-relevant constructs expressed less prejudice and were more willingness to empower immigrants. In addition, the effects were not moderated by individual differences in prejudice-relevant variables (i.e., Disgust Sensitivity, Intergroup Disgust-Sensitivity, Intergroup Anxiety, Social Dominance Orientation, Right-wing Authoritarianism). Study 2 considered a different target category (i.e., Blacks) and attempted to strengthen the effects found by comparing the impact of subliminal empathy primes (relative to no prime or subliminal primes of empathy paired with Blacks) on explicit prejudice towards marginalized groups and Blacks, willingness to help marginalized groups and Blacks, as well as implicit prejudice towards Blacks. In addition, Study 2 considered potential mechanisms for the predicted effects; specifically, general empathy, affective empathy towards Blacks, cognitive empathy towards Blacks, positive mood, and negative mood. Unfortunately, using subliminal empathy primes “backfired”, such that exposure to subliminal empathy primes (relative to no prime) heightened prejudice towards marginalized groups and Blacks, and led to stronger expectations that marginalized groups and Blacks improve their own situation. However, exposure to subliminal primes pairing empathy with Blacks (relative to subliminal empathy primes alone) resulted in less prejudice towards marginalized groups and more willingness to directly assist Blacks, as expected. Interestingly, exposure to subliminal primes of empathy paired with Blacks (vs. empathy alone) resulted in more pro-White bias on the implicit prejudice measure. Study 2 did not find that the potential mediators measured explained the effects found. Overall, the results of the present project do not provide strong support for the use of subliminal empathy primes for improving intergroup relations. In fact, the results of Study 2 suggest that the use of subliminal empathy primes may even backfire. The implications for intergroup research on empathy and priming procedures generally are discussed.

Individual Differences in Global/Local Processing Bias and the Attentional Blink

When the second of two targets (T2) is presented temporally close to the first target (T1) in rapid serial visual presentation, accuracy to detect/identify T2 is markedly reduced as compared to longer target separations. This is known as the attentional blink (AB), and is thought to reflect a limitation of selective attention. While most individuals show an AB, research has demonstrated that individuals are variously susceptible to this effect. To explain these differences, Dale and Arnell (2010) examined whether dispositional differences in attentional breadth, as measured by the Navon letter task, could predict individual AB magnitude. They found that individuals who showed a natural bias toward the broad, global level of Navon letter stimuli were less susceptible to the AB as compared to individuals who showed a natural bias toward the detailed, local aspects of Navon letter stimuli. This suggests that individuals who naturally broaden their attention can overcome the AB. However, it was unclear how stable these individual differences were over time, and whether a variety of global/local tasks could predict AB performance. As such, the purpose of this dissertation was to investigate, through four empirical studies, the nature of individual differences in both global/local bias and the AB, and how these differences in attentional breadth can modulate AB performance. Study 1 was designed to examine the stability of dispositional global/local biases over time, as well as the relationships among three different global/local processing measures. Study 2 examined the stability of individual differences in the AB, as well as the relationship among two distinct AB tasks. Study 3 examined whether the three distinct global/local tasks used in Study 1 could predict performance on the two AB tasks from Study 2. Finally, Study 4 explored whether individual differences in global/local bias could be manipulated by exposing participants to high/low spatial frequencies and Navon stimuli. In Study 1, I showed that dispositional differences in global/local bias were reliable over a period of at least a week, demonstrating that these individual biases may be trait-like. However, the three tasks that purportedly measure global/local bias were unrelated to each other, suggesting that they measure unique aspects of global/local processing. In Study 2, I found that individual variation in AB performance was also reliable over a period of at least a week, and that the two AB task versions were correlated. Study 3 showed that dispositional global/local biases, as measured by the three tasks from Study 1, predicted AB magnitude, such that individuals who were naturally globally biased had smaller ABs. Finally, in Study 4 I demonstrated that these dispositional global/local biases are resistant to both spatial frequency and Navon letter manipulations, indicating that these differences are robust and intractable. Overall, the results of the four studies in this dissertation help clarify the role of individual differences in attentional breadth in selective attention.