Affective responses to non-sexual imagery depicting gay men and lesbian women

ABSTRACT

When examining negative attitudes and behaviors directed toward gay men and lesbian women (i.e., homonegativity), researchers tend to use measures that require participants to respond to belief statements. This methodology is problematic for two reasons: 1) it focuses on the social categories “gay men” and “lesbian women” and ignores the practices of relational intimacy engaged in by gay and lesbian persons (practices that, arguably, are at the crux of homonegativity); and 2) it overlooks the affective responses that sexual minorities evoke in heterosexual people. These issues were tackled in the current study. Specifically, heterosexual participants (N = 241) were asked to report their affective state using six basic emotions while viewing photos depicting male-male, female-female, and heterosexual couples. Findings demonstrated that participants, regardless of gender, reacted most negatively to images of female-female couples engaging in everyday intimacies. Theoretical explanations for these findings are explored.

KEYWORDS:

• Affective response
• gay men
• lesbian women
• homonegativity
• everyday intimacies
• symbolic threats

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The data described in this article are openly available in the Open Science Framework at 10.17605/OSF.IO/4NKHD

Open Scholarship

This article has earned the Center for Open science badges for Open Materials through Open Practices Disclosure. The materials are openly accessible at 10.17605/OSF.IO/4NKHD

Notes

1. Based on the results of Oppenheimer, Meyvis, and Davidenko (2009) we opted to include an instructional manipulation check (IMC) item to determine how many participants were actually paying attention to the requirements of their participation versus merely satisficing. The IMC involved embedding an item that asked participants to provide a specific response in the instructions. Had participants read the instructions, they would have easily recognized how to correctly respond. For those who were providing arbitrary responses, they would have responded incorrectly. While a 25% exclusion of cases may seem high, it is actually a relatively low amount of data loss since Oppenheimer et al. reported that 46% of their sample failed to answer their IMC question correctly.

2. Post-hoc analyses were carried out assessing the mean attractiveness for each of the six groups of images. See Table 4. The symbolic threats – lesbian women group was rated overall as statistically significantly more attractive than the everyday intimacies – gay men (t [18] = 2.50, p = .022), everyday intimacies – lesbian women (t [18] = 3.10, p = .007), and everyday intimacies – heterosexual couples (t [18] = 2.51, p = .022) groups. In addition, the symbolic threats – heterosexual couples images were rated as statistically significantly more attractive than the everyday intimacies – lesbian women group (t [19] = 2.96, p = .008). No other group differences emerged. See Table 4.

3. For details on the measures used and sample images, visit https://osf.io/4nkhd/?view_only=46586eca34c640d0a6f3c3017c89972c.

4. Resources regarding the appropriate treatment of anger on the basis of multicollinearity could not be located (i.e., should it be removed from the model altogether or retained as a covariate given its more modest correlations with the other dependent variables: surprise, fear, and sadness?). Since the same pattern of statistically significant results emerged in both models, we decided to report the model with anger removed.

5. Happiness was not included as a dependent variable because it did not conceptually link with the other affective states (i.e., Cronbach’s alpha coefficient for the six affective states [i.e., happiness, anger, disgust, surprise, fear, and sadness] increased from .35 to .83, when happiness was removed).

6. While it may seem counterintuitive to conduct a univariate test to examine the group differences in a multivariate function (indeed, this is why descriptive discriminant analysis [DDA] is a more desirable form of multivariate post hoc testing), it is important to remember that the discriminant scores entered into the analysis were derived from including each of the four independent variables into the DDA. Therefore, the results of a multivariate analysis are being tested. Note that the group centroids are the means of the discriminant scores for each category of the dependent variable. Applying post hoc testing in this manner allows for between-groups testing across the group centroids which assists in the interpretation of group differences via 95% confidence intervals and ensuring that researchers do not need to rely upon “eye-balling” the outputs.

7. When more liberal inclusion criteria were used (i.e., cases that failed the IMC were retained and sexual orientation was assessed solely using self-identification) but multivariate outliers were still removed, the statistically significant results were identical. We feel that similar outcomes emerging when using maximally stringent inclusion criteria further strengthen our results while simultaneously minimizing the concern of Type II error. Therefore, we stand by our decision to exclude the cases that failed to correctly answer the IMC.

8. Will & Grace and Orange is the New Black were selected as examples because they depict fictional characters that can realistically exist in the world today. Other popular television series such as Game of Thrones and Once Upon a Time which also depict gay and lesbian characters (e.g., Renly Baratheon and Red Riding Hood) are more fantastical and, possibly, less likely to constitute meaningful instances of parasocial contact.

Additional information

Funding

This work was supported by the Social Sciences and Humanities Research Council of Canada.

Notes on contributors

CJ Bishop

CJ Bishop is a postdoctoral research associate at the University of Alberta. CJ’s research interests include prejudice and discrimination toward sexual and gender minority (SGM) groups, psychometrics, and other aspects of human sexuality.

Todd G. Morrison

Todd G. Morrison is a Professor in Social Psychology at the University of Saskatchewan. Todd conducts research in the areas of sexual and gender minority psychology, gay pornography, body image, and feminism.