An overwhelming number of adolescents (approximately 60%) involved in the juvenile justice system meet Idiagnostic criteria for one or more psychiatric disorders. This poses a substantial burden for state juvenile justice systems that generally have some responsibility to care for the minors in their custody. Unfortunately, interpretation of prevalence rates for childhood mental disorders is not straightforward due to unreliable testing methods and hinderances to defining childhood abnormalities. Complications in the identification of childhood mental disorders include symptom overlap in current nosological frameworks, changes in the expression of symptoms across developmental periods, and inconsistencies between genders and across demographic groups (e.g., Hispanics vs. Whites). The proposed study was designed to improve mental health screening procedures for juvenile justice facilities, and advance understanding of juvenile psychopathology, by analyzing the perfomance of a juvenile justice screening tool using innovative test construction methods. Just over 36,000 cases from the National Norm data for the Massachusetts Youth Screening Instrument, Second Version (MAYSI-2) and item response theory will be used to (1) validate the MAYSI-2's structural validity and measurement of mental, emotional, and behavioral disturbances; (2) investigate gender and ethnic/racial-specific differences in the measurement or expression of these disturbances; and (3) investigate age-related differences in the measurement and expression of these disturbances. IRT tests for measurement bias will distinguish test items that may be seen as consistent across gender, ethnic/racial, and age groups; from symptoms that should be considered racially, developmentally, and/or gender-specific. Findings will have the immediate benefit of enhancing the utility of the MAYSI-2 by estimating group-specific cut-off scores. Indirectly, findings will work toward advancing our misunderstanding of the nature and development of child psychopathology in general.
