Joel Anderson

Date of Submission



Implicit anti-gay attitudes are relatively unconscious, automatic evaluations of gay men and lesbians which are measured by assessing the strength of associations in a speeded classification task. In contrast to other implicit prejudices (e.g., racism, sexism) there are unique challenges to overcome when measuring implicit attitudes toward gay men and lesbians. For example, there is no visible characteristic that can be reliably used to identify sexual orientation, nor are there any names, and only a few nouns (e.g., gay) which are uniquely associated with this social category. As the measurement of implicit anti-gay attitudes relies on the presentation of at least six stimuli to represent the social category, continued discussion on stimuli selection is needed. To date, researchers have relied on the use of stimuli that are conceptually related to the category of GAY (e.g., the rainbow flag, same-sex wedding cake toppers), and therefore these measures may be eliciting related attitudes (e.g., attitudes toward the amorphous category GAY, rather than to GAY PEOPLE). The main aim of this thesis was to present a new approach which addresses this shortcoming. I provided evidence for the person-based approach to implicit antigay attitudes in two initial studies. Study 1 demonstrated that presenting faces of straight male, straight female, gay male, and lesbian target stimuli (who are known for their sexual orientation) with opposite gender distracter stimuli elicits implicit gender attitudes consistent with previous research (Rudman & Goodwin, 2004). However, the same set of gay target stimuli presented with straight distracter stimuli of the same gender (e.g., lesbian targets, and straight female distracters), substantially reduced and reversed the pattern of results, such that gay men are weakly implicitly associated with positive and lesbians are weakly implicitly associated with negative. Moreover, these patterns are affected by participant’s own gender and sexual orientation (Study 2). These findings are interpreted as evidence that the person-based approach is assessing constructs of implicit gender attitudes and implicit sexual orientation-based attitudes that are distinct. Furthermore, Study 3 replicated the results of previous implicit prejudice research (using stimuli that have typically represented gay men and lesbians in implicit measures; e.g., Nosek, 2005) and the findings of Study 1 (i.e., using the person-based approach) providing evidence of the meaningful differences between implicit attitudes towards the category GAY and to GAY PEOPLE. Finally, studies 4 and 5 explored the role of religion and religiosity, known predictors of anti-gay attitudes, on implicit person-based antigay attitudes. Study 4 revealed that only religious fundamentalism was a strong predictor of explicit gay attitudes, and that no significant regression model was found that predicted implicit person-based anti-gay attitudes. In contrast, Study 5 used contextual variation to prime the construct of religion (i.e., distracter stimuli were faces of religious individuals, such as nuns and priests) and revealed that relevant religious stimuli led to a subsequent increase in positive implicit person-based attitudes towards gay people for Atheist, but not Christian participants. Taken together, these findings provide strong evidence for the person-based approach to anti-gay attitudes, suggest that implicit prejudice towards gay people differs from in important ways from implicit attitudes towards the category gay, and demonstrated that implicit person-based anti-gay attitudes have a meaningful, but complex relationship with religiosity. As a result, the proposed measure of implicit person-based anti-gay attitudes makes a novel and important contribution to the current anti-gay literature and provides researchers with a much needed and well validated alternative to the typical approach.


School of Psychology

Document Type


Access Rights

Open Access


252 pages

Degree Name

Doctor of Philosophy (PhD)


Faculty of Health Sciences

Included in

Psychology Commons