Abstract
Several studies have investigated the predictive validity of the Hare Psychopathy Checklists (PCL); however, the majority have used North American samples. The PCL can inform important decisions about people’s futures and therefore it is imperative to validate its psychometric properties within the jurisdictions it is used, particularly for Indigenous and non-European populations. In New Zealand, Botha and Polaschek (2023) concluded that the PCL: Screening Version (PCL:SV; Hart et al., 1999) demonstrated cross-ethnic predictive invariance using traditional discrimination statistics. The present study extends on this research by reporting on a novel statistic for assessing fairness in risk tool performance, cross-area under the curve (xAUC), as well as error rate balance and predictive parity of the PCL:SV for the prediction of any and violent 5-year recidivism outcomes in European and Māori subsamples (n = 288). In contrast to previous research, we found discrepancies, between European and Māori, on xAUC, error rate balance, and predictive parity.
Get full access to this article
View all access options for this article.
