Scale-level factor analysis of the MACI and MMPI-A with a court-ordered adolescent sample

The current study examined the scale-level factor structure and canonical relationship of two widely used measures of personality and psychopathology, the Minnesota Multiphasic Personality Inventory – Adolescent (MMPI-A) and the Millon Adolescent Clinical Inventory (MACI), with a mixed gender court-referred adolescent sample. Previous research has suggested factors derived from scale-level factor analytic studies of the MMPI-A and MACI are useful as they provide information about the structure and organization of the tools and the derived factors can be used as an alternative interpretive approach. Previous factor analytic studies of the MACI have found largely disparate results; therefore, the current study used confirmatory factor analysis (CFA) in an attempt to replicate the most recent empirical factor-structure of the MACI (Newman et al., 2015). Although the CFA of the current sample did not exactly fit the factor structure identified by Newman et al. (2015), follow up exploratory factor analysis yielded a two factor solution which was close to the Newman et al. (2015) findings. The current study lends support to the idea the Personality Pattern and Clinical Syndrome scales of the MACI can be collapsed into broad Externalizing and Internalizing factors for interpretation for a general court-referred sample. Previous factor analytic studies of the MMPI-A found fairly consistent results and the current study expanded upon this research by replicating the factor structure previously identified in the MMPI-A Structural Summary using confirmatory factor analysis with a mixed gender sample from an array of court-referred placements while previous research has largely focused on males from community, psychiatric, and secure detention settings. The current study supports the use of the MMPI-A Structural Summary with a court-referred sample containing both males and females. Results of canonical correlation analyses suggested a high degree of shared variance between the MACI and MMPI-A for a court-referred sample; therefore, these measures may be somewhat redundant measures for this population.Department of Educational PsychologyThesis (Ph. D.