Abstract
Previous research has suggested that scores generated using a sexual orientation Implicit Relational Assessment Procedure (IRAP) correlate with self-reported sexual orientation in gay and heterosexual men and have a strong ability to differentiate the two groups (Timmins et al., 2016). The present study sought to replicate this work and to determine how a range of groups that differ in terms of self-reported sexual orientation perform on the measure. Lesbian/gay, heterosexual, and bisexual women (n = 99) and men (n = 96) completed a sexual orientation IRAP and self-report measures of sexual attraction and behavior. The IRAP involved responding "True" or "False" to pictures of nude males and females paired with words meaning attractive or unattractive. Participants were required to respond as if all men were attractive and all women were unattractive for half of the IRAP's test blocks, and vice versa for the other half. Response latencies were recorded, and D-IRAP scores were calculated for overall responses, female responses, and male responses. Significant correlations were found between all D-IRAP scores and all corresponding self-report measures for men (rs = .38-.64). This was also true for women (rs = .24-.53) with a single exception: the D-IRAP score for responses to men with self-reported sexual attraction to men. Similarly, almost all D-IRAP scores significantly differentiated sexual orientation groups (areas under the curve = .64-.92), apart from bisexual and lesbian/gay women. These findings suggest the IRAP is a useful ex situ measure of sexual orientation.