Files
Abstract
The purpose of this dissertation is to provide evidence for instrument validation through the assessment of psychometric properties. Specifically, it will demonstrate methods for validating the structural aspect of validity on two scales intended to provide measurement of psychological constructs within the juvenile offender population. To do this, two separate studies were conducted. In study one, factor structure of the Burns Brief Mood Survey (BMS) was tested through the use of exploratory factor analysis. For study two, confirmatory factor analysis was used to assess the hypothesized factor structure of the Juvenile Offender Parent Questionnaire (JOPQ). In study one, data analysis yielded a simple five factor structure for the BMS; however, this factor structure differed from the hypothesized factor loadings. These results bring to light the importance of structural validation when utilizing an instrument on a specialized population. Results of study two confirmed the hypothesized six factor structure of the JOPQ. The model demonstrated an adequate level of fit, but item level analysis and examination of the modification indices suggested changes to the structure of the instrument that may improve model fit.