|
Abstract
|
Despite considerable research on the predictive accuracy of risk scales, there is limited research exploring other factors that influence perceptions of risk. We recruited participants (N = 1,955) from Amazon's Mechanical Turk to read a vignette about a fictional intimate partner violence offender, varying risk level on a fictional scale (low or high), perpetrator gender (cis male, cis female, or transgender female), victim gender (cis male or cis female), and mental health diagnosis (none, schizophrenia, bipolar disorder, or anxiety disorder). The strongest effect was for the risk scale, with offenders perceived as highest risk when the scale reported "high risk" as opposed to "low risk." The other main effects were also statistically significant. Cases were perceived as riskier when the perpetrator was cis male or the victim was cis female. Regarding mental health diagnosis, the highest risk ratings were provided in the schizophrenia condition. There was also a significant interaction among risk level, perpetrator gender, and diagnosis. The extent to which participants relied on gendered stereotypes about the relationship between mental illness and violence when providing risk judgments should be examined in future research. Overall, these findings enhance our understanding of characteristics that are secondary to risk level but are likely to influence case management decisions in cases of intimate partner violence.
Language: en |