(PS12-C60) Understanding Potential Perpetrator Behavior: The Development of the Perceptive Evaluation of Rape Proclivity (PERP) Scale

The behavioral intentions contributing to sexual violence (SV) are known as rape proclivity (RP) and can be defined as an individual’s endorsement of the likelihood to commit SV. Given the importance of adequately assessing RP to address and reduce SV, Malamuth (1981) developed the Attraction to Sexual Aggression (ASA) scale, which provides 14 activities and the following instructions: If you could be assured that no one would know and that you could in no way be punished for it, how likely would you be to do the following? Of the 14 items, “Rape” and “Force a female to do something sexual she didn’t want to do” are the only 2 related to SV. Unfortunately, terms like “rape” and “force” may be too reactive and prone to socially desirable responding, especially given evolving values for acceptable language and behavior.

Updating traditional scales that measure sensitive topics has been productive in several areas (e.g., sexism, racism) and may be necessary to facilitate empirical investigation of RP. Considering sociopolitical changes within the 40 years since RP’s introduction, measurement that assesses intent to perpetrate in less overt/socially reactive ways and that includes more variants of misconduct is arguably warranted. Therefore, 14 items assessing varying degrees of SV were developed and interposed with 14 distractor items related to other crimes – yielding the PERP. Target items span 4 categories of perpetrator behavior: coercion (I would try and get someone to give me oral sex even if it took a lot of persuasion), disregard for consent (I would still sleep with someone even if they’re a little too drunk, high, or otherwise out of it), verbal/physical threats (I would use my power and dominance to get someone to have sex with me), and hostile/overt (I would use my size/strength to hold someone down in order to have sex with them). Target items aim to assess behavioral intent while increasing the severity of perpetration behaviors, with the final question closely replicating Malamuth’s: I would rape someone if I knew could not get caught or punished.

The 28-item PERP was administered to 306 cis-gender male MTurk workers. Responses were dichotomized to reflect what is typically done in this area (i.e., responses at or above 2 on a 5-point scale are classified as Some Likelihood to Perpetrate); however, the PERP was made stricter such that responses at or above a 3 were classified as Moderate Likelihood to Perpetrate. The PERP identified 217 participants (70.9%) with a moderate likelihood to perpetrate compared to the established measure which identified 155 (50.7%) using a more lenient dichotomy. The PERP also identified 67 more participants who indicated moderate or higher likelihood to perpetrate that went undetected by the ASA. In contrast, the ASA captured only 5 more respondents with some likelihood to perpetrate that were missed by the PERP. This suggests that the PERP may detect more nuanced potential for perpetration (e.g., coercion, use of threats) than the ASA (x² = 101.84, p < .001). The PERP demonstrated excellent internal consistency (α = .98), convergent validity (ASA, r = .53; IRMA-2011, r = .64), and discriminant validity (BIDR-6: Psychopathy, r = -.20). Additional analyses (e.g., EFAs) will be presented and discussed.