The relationship between clinical improvement and client perceived helpfulness in an acute partial hospitalization program : a canonical correlation analysis

No Thumbnail Available
Authors
Morrison, Dennis P.
Advisor
Duckworth, Jane C.
Issue Date
1985
Keyword
Degree
Thesis (Ph. D.)
Department
Other Identifiers
Abstract

Specialized partial hospitalization programs have been used to treat a variety of mental health problems however, there has been little research done on the relationship between clinical improvement and perceived helpfulness in such programs. In this study, improvement and perceived helpfulness data collected on 923 clients who attended a partial hospitalization program for acutely disturbed improvement data set consisted of intake and discharge scores on eight scales. Three scales provided affective improvement data, three provided problem resolution information, one provided goal attainment information, and one measured improvement in functionality. Preliminary results indicated that three of the scales would have to be omitted from further analysis. The perceived helpfulness data set was made up of the clients' weekly ratings of helpfulness of 19 therapeutic experience in the partial hospitalization program. These two sets were analyzed using a canonical correlation analysis and were found to be significantly correlated (p<.001). Based on the-respective canonical variate loadings, the improvement composite was interpreted as global improvement and the perceived helpfulness composite was interpreted as general perceived helpfulness. It was viewed as somewhat surprising that no clear pattern of perceived helpfulness emerged nor wasindividuals were analyzed in a post hoc fashion. The improvement clearly of one type. A second analysis was performed to assess the ability of the Day Therapy Appropriateness Scale (DTAS) to predict improvement in therapy in an acute partial hospitalization program. A stepwise multiple regression analysis was performed on data from 304 clients to see if there was a significant (a<.01) relationship between their scores on the DTAS and a linear composite of a set of five improvement variables. The result of this analysis was not significant which supports work previously done with the DTAS.