THE UNIVERSITY OF TENNESSEE, 1983

                         PSYCHOLOGY, CLINICAL (0622)
 

                         The present study was designed to examine the relationship between network variables, preparation for
                         therapy, and therapy attendance. Clients presenting for intake at a mental health center were randomly
                         assigned to treatment and control groups. Subjects completed an expectations questionnaire. Subjects
                         were seen for intake and then interviewed by the author about their network. Half were given therapy
                         preparation. It has been hypothesized that prepared subjects would show changes in expectations; this
                         held only for a decrease in expectation of expert advice. Analysis of change scores revealed no
                         difference between prepared and nonprepared clients. A multiple regression of network variables, the
                         preparation condition, gender, and score on social desirability, against the number of sessions attended
                         was conducted. The model could account for 14% of the variance. Low scores on social desirability,
                         network available for the discussion of personal matters, and provision of feedback were positively
                         associated with attendance. The index of network for fun and relaxation, and network involved in the
                         provision of material aid, were negatively associated with attendance. A second regression of same
                         variables against classification by attendance at one session vs. more was conducted. This yielded a
                         significant model, accounting for 18% of the variance. Again, low score on social desirability, network
                         available for discussion of private feelings, and network involved in support for treatment were associated
                         with attendance. A discriminant function was conducted. The model classified 71% of the subjects
                         correctly. It was hypothesized that preparation would be associated with attendance. The relationship
                         was positive but nonsignificant. A positive relationship between attendance and network was expected.
                         Results were mixed, but findings point to the importance of having people to provide feedback and
                         discuss private feelings. Social desirability scores suggest a nondefensive approach is important for
                         treatment involvement. Analysis revealed nonsignificant findings between ratings of network need and
                         satisfaction with attendance. Women were found to have smaller networks across all functional
                         categories. Differences in network size were also found in association with marital status. The findings
                         lend credence to the predictive value of network variables and suggests need for investigation into
                         particular functions and specific preparation techniques.

Social Systems Simulation Group P.O. Box 6904 San Diego, CA  92166-0904 Roland Werner, Principal Phone/FAX  (619) 660-1603