Can we measure subtle discrimination of homosexuals? The effects of learning persons' sexual orientation and appearance on interaction intentions

Abstract

Although members of LGBT community are often faced with overt discrimination, they report being even more frequently exposed to subtle discrimination – non-obvious, often unintentional, however biased treatment. Due to its subtle nature, it is difficult to detect. One of its forms might be presuming gender-atypical appearance (feminine males or masculine females) to be a predictor of homosexual orientation, as well as the behaviour that emerges from this biased presumption. Having in mind that deviations from male gender roles and typical appearance are more severely sanctioned than deviations from female roles, males could be more exposed to this type of bias. In the current research, we focused on how (a) gender-atypical appearance and (b) information about sexual orientation (SO) affect the willingness to cooperate with the persons concerned. To test this, we devised a within-subjects design – 2 (appearance typicality: masculine/feminine) x 2 (sexual orientation: hetero/homosexual), with interaction intention as a dependent variable assessed before and after exposing SO. Gender-typicality of appearance was operationalized via photographs depicting the stimulus-persons in typically masculine or feminine sitting postures, while their SO was provided via short profiles. Firstly, the participants’ task was to specify the extent to which they were willing to cooperate with the stimuli-persons regarding their physical appearance, and afterwards to provide the same estimation by combining the information about their SO. Willingness for cooperation was assessed on the 7-point Likert scale. Drawing from a convenient sample of 41 heterosexual men, aged 19-47 (M=21.463, SD=4.382), we demonstrated that subjects initially preferred cooperation with people depicted in masculine posture as opposed to those in feminine posture (M(m)=4.213, SD(m)=0.914; M(f)=3.567, SD(f)=1.095; F(1,40)=16.106, p=.000, η2=.287). As expected, after being provided with information about stimuli person’s SO, subjects increased ratings of those who were pictured in feminine posture but labelled as heterosexuals (M1=3.567, SD1=1.095; M2=4.232, SD2=1.230; F(1,40)=14.179, p=.000, η2=.262). We interpreted this change as a subtle way of discriminating homosexuals – or to put it differently, of favouring heterosexuals. Their robustness needs to be further backed up with evidence, however, the paradigm we designed seems suitable to detect this type of bias.