| Abstract: | This paper raises and attempts to resolve some significant and controversial issues in the design of family therapy outcome studies. First, the issue of clinical relevance is addressed. Much of psychotherapy research has little relevance for practicing clinicians; this paper discusses some partial explanations and proposes some changes in the way data are reported that would have a salutary effect. Second, inter-model comparative outcome studies are critiqued. Third, the necessary preconditions for internally valid outcome studies are delineated, along with conditions in which randomization and the use of control groups can be abandoned. Fourth, there is a discussion of a research strategy which examines the generalizability of treatment effects found in research settings to conditions which come closer to resembling clinical practice. Fifth, questions regarding the selection of therapists for family therapy outcome research are addressed, including the issue of optimal level of therapist experience, and the question of whether or not “therapists” should be treated as a randomized factor in the overall design. Finally, there is a brief discussion of outcome criteria, with a fairly clear-cut and straightforward recommendation for determining whether or not a family therapy has had a desirable effect. |
