Self-report and performance-based measures of impulsivity in justice-involved youths: Predictive validity and dynamic change

Impulsivity is a multi-faceted construct comprised of lack of premeditation, urgency related to negative affect, sensation seeking and lack of perseverance with difficult tasks (Whiteside & Lynam, 2001). Impulsivity is also a key risk factor for recidivism (Andrews & Bonta, 2010), and particularly relevant to adolescent behaviour (Hirschi & Gottfredson, 1983; Steinberg, 2007). An understanding of dynamic risk factors that are amenable to change, such as impulsivity, is important for risk management and intervention (Douglas & Skeem, 2005). Prior research has typically relied upon self-report questionnaires, but performance-based measures that focus on completing a specific task may provide unique information on impulse control (Reynolds, Penfold, & Patak, 2008), and conduct problems (Mathias, Marsh-Richard, & Dougherty, 2008). Minimal research has compared different measures of impulsivity in adolescent populations, or examined whether dynamic change relates to outcomes. The purpose of this study was to examine whether self-report (Eysenck Junior Impulsiveness Scale; EJIS; Eysenck, Easting & Pearson, 1984) and performance-based (GoStop Impulsivity Paradigm; GoStop; Dougherty, Mathias, & Marsh, 2003b) measures of impulsivity related to each other, changed over time, and whether change differentially predicted reoffending and violence. Participants were 107 adolescents (35 females) between ages 12 and 18 who were supervised on community probation orders. Youths completed questionnaires during semi-structured interviews with a 6-month interval. Recidivism was measured through self-report and official records. Results indicated that the two impulsivity measures were not associated, but both predicted prospective outcomes of reoffending, aggression and violence. GoStop scores significantly contributed to the prediction of recidivism, even after controlling for EJIS and age of first arrest, supporting the notion that performance-based impulsivity measures provide complementary information to self-report. Group-level changes were non-significant, but intra-individual changes were observed for a small group of youths over the 6-month follow up period. Lower reported physical victimization, substance use, and negative affective states, including anger and depression/anxiety, were associated with gains in impulse control. Change scores predicted self-report and official records of general and violent reoffending at 6- and 12-month follow ups. Thus, both self-reported and performance-based impulsivity are relevant to outcomes of reoffending and violence among justice-involved youths.