The working alliance: rehabilitation outcomes for persons with severe mental illness

Journal of Rehabilitation,  April-June, 2004  by Chandra M. Donnell,  Daniel C. Lustig,  David R. Strauser

• E-mail
• Print
• Link

Limitations

Conclusions about the results are limited by the following considerations. First, this study utilized an ex post facto design. A limitation of ex post facto designs is the difficulty determining a causal link between variables. Other factors may have affected the outcome. For example, clients may have refused service when the initial counselor contact was problematic in terms of the development of a working alliance. Thus, there may have been some pre-selection of client and counselor. Other variables that could affect the outcome include family and financial support, training of the counselor, and the duration of service. Second, the BBER/CMS was unable to contact slightly less than half of the potential respondents (47%) or were contacted but refused to reply (7%) and it is unclear whether non-respondents differ significantly from respondents. Third, interviews were completed during the 1999-2000 fiscal year with Tennessee Division of Rehabilitation Services clients. Consequently, the interpretation of the results should be limited to the sample examined at the time of the study. Fourth, although care was taken to provide evidence of the reliability and validity of the measure of working alliance, more evidence is needed to substantiate the reliability and validity of the Working Alliance Survey. Finally, only the client's view of the strength of the working alliance was used to measure working alliance. While research provides evidence that client ratings of the alliance are stronger predictors of treatment outcomes than counselor ratings (Connors et al., 1997; Horvath & Symonds, 1991; Luborsky, 1994), the validity of the measure of working alliance may be increased if both counselor and client were asked their perception of the working alliance.

Future Research

Given the results of this study, there are several recommendations to be made regarding future research in this area. This study developed a measure of working alliance (WAS) based on existing survey questions, not originally designed to address the working alliance construct. However, a number of measures have been previously developed (Horvath, 1994, and Horvath & Greenberg, 1989). It would be important to replicate this study utilizing instruments that have previously been developed and have been widely used in research on the working alliance.

Future replication of this study may also benefit from administering the working alliance measure at various time periods throughout the counseling interaction. Previous research (Krupnik et al., 1996) has indicated that the association and magnitude of outcome variance explained was considerably stronger when utilizing a mean alliance score. Administering the measures at various intervals during the therapeutic process, and examining the mean score, may allow researchers a more comprehensive understanding of the working alliance versus an examination at the conclusion of services, which may be confounded by client satisfaction with outcomes. Similarly, replication of this study may also benefit from an examination of the counselors' views of the working alliance. Counselors and clients often have divergent views on the therapeutic process and treatments, which can impact perception of the working alliance. Input from the counselors may be pertinent to determining the formation and strength of the therapeutic relationship (Chinman, Rosenheck & Lam (2000).