Effects of therapist self-disclosure on client perceptions of the therapeutic alliance and session impact

It has been shown that different types of therapist self-disclosure (TSD) positively affect client perceptions of therapist characteristics such as expertness, attractiveness, and warmth. This study broadened the focus of client perceptions to more global therapy process variables, the therapeutic alliance and session impact. It was hypothesized that positive client perceptions of therapist characteristics would be related to and would influence perceptions of session-level process variables but that differential effects would be found for two groups of subjects, African American and Caucasian. Subjects participated in a 40-minute life-history interview and were exposed to one of three levels of disclosure: no TSD; factual demographic information; or personal information about past or present experiences and feelings. Using MANOVA, no significant differences between dependent measures of therapist characteristics, therapeutic alliance dimensions, and session impact dimensions were found for the main effects of ethnic group and level of disclosure or for the interaction of these variables. A path model specifying direct and indirect relationships between these variables was not supported nor was a revised path model with nonsignificant paths deleted and others added. These findings resulted from a large ceiling effect in which the interviewer and the session were rated very highly across all groups and levels of TSD. The lack of variability across ratings made it nearly impossible to obtain significant results. It was found that while TSD was not significantly correlated with any of the dependent variables, ratings of therapist characteristics, alliance dimensions, and session impact were all highly correlated. Examination of partial correlations between variables demonstrated a potential mediating effect of therapist characteristics on the relationship between TSD and alliance dimensions and a mediating effect of alliance dimensions between therapist characteristics and session impact. Thus while the results of this study are statistically nonsignificant, they demonstrate that participant perceptions of a therapy session radically differ from those types of ratings made by an observer of the session. This highlights the need for future TSD research to examine the phenomenon using live interviews and to determine which aspects of the therapist and session lead to various participant ratings.