An investigation of gender bias in a juvenile justice setting.

Accurately identifying mental health needs of adolescents in juvenile justice settings is crucial to ensure treatment needs are met (Baum et al., 2009; Grisso, 2005). One instrument that can be used to identify mental health needs of adolescents is the Minnesota Multiphasic Personality Inventory – Adolescent (MMPI-A; Butcher et al, 1992), a measure widely used in research and practice (Archer & Newsome, 2000). Ethical and legal guidelines for assessments in juvenile justice settings require the use of tests with known psychometric properties, including evidence supporting cultural fairness of scale scores. To date, most research investigating the psychometric properties of MMPI-A scores in juvenile justice settings has been conducted using exclusively male samples (Baum et al., 2009, Veltri et al., 2009) with only one study examining differences in these properties across boys and girls (Handel et al., 2011). As such, the current study used a sample of 244 (161 males, 83 females) adolescents court ordered to receive a psychological evaluation to investigate two research aims. These included: 1) to provide validation evidence for scores on MMPI-A Content scales for boys and girls and 2) to investigate predictive bias by gender for MMPI-A Content scale scores. Generally, results from Pearson’s r correlations provided evidence supporting the convergent and discriminant validity of MMPI-A content scale scores in boys and girls (Convergent: r’s = .30 to -.57, p < .001; Discriminant: r’s < .30, p < .001). Additionally, results of regression analyses examining gender as a moderator demonstrated that gender did not significantly influence the relation between MMPI-A Content scale scores and BASC-2 scores. Overall, these findings support the validity of scores on MMPIA content scales in boys and girls and provide evidence suggesting that predictions made by MMPI-A content scale scores in juvenile justice settings may be culturally fair with respect to gender.