No abstract

Twenty master’s level clinical functions and training models were either endorsed, or not endorsed by chief psychologists from 396 mental health agencies. The 396 cases were divided into two halves with each half randomly selected to provide proportional representation from each facility type, and the two halves were separately factor analyzed. Several methods of rotation were examined. The method selected was orthogonal since it most reliably reproduced the same factors in both samples. Four factors were retained after being found reliably reproduced across both samples. The entire sample was then factored, and factor scores for each reliable factor were obtained for each case. The quality of the factor scores was checked by computing the intercorrelations of the factor scores and comparing them to the expected intercorrelation of factors. Mean factor scores were generated for each of nine mental health agency types and displayed graphically. Planned comparisons across agency types were made for mean factor scores for three of the four reliable factors retained. All comparisons were significant and differences were in the expected direction.