15-04-2021
Measurement Equivalence across Multiple Administrations of the Massachusetts Youth Screening Instrument
Gepubliceerd in: Journal of Psychopathology and Behavioral Assessment | Uitgave 4/2021
Log in om toegang te krijgenAbstract
Multiple studies have shown that the Massachusetts Youth Screening Instrument-2nd Version (MAYSI-2) has strong validity and reliability. Justice involved youth are often administered the MAYSI-2 on repeated occasions. The purpose of the current study was to test the equivalence of measurement properties over time for males who were administered the MAYSI-2 on multiple occasions. Archived data were drawn from the Utah Juvenile Justice Services administration, which consisted of all administrations of the MAYSI-2 between 2007 to 2018. Among these, 909 males completed the MAYSI-2 three or more times. Single and multiple group longitudinal invariance tests were conducted using a weighted least squares mean and variance adjusted estimation in order to determine measurement equivalence across multiple administrations. Increasingly restrictive constraints for configural, metric, and scalar invariance models across Time 1 and Time 3 as well as models constraining latent means and variances to be equal across Time 1 and Time 3 all fit these data well. The established seven factor structure for the MAYSI-2 fit the data as well. However, some differences were noted in item level functioning across multiple administrations. The results of the current study suggest that multiple administrations of the MAYSI-2 may not exhibit equivalent measurement properties over time. Consistency of the measurement properties of the MAYSI-2 over time is essential for evaluating changes in mental health. Given that contact with the juvenile justice system has the potential to worsen symptoms of mental illness, caution is warranted when using the MAYSI-2 to assess change.