ABSTRACT
This chapter explains how psychotheraphy research can benefit from the application of multilevel modeling (MLM). It illustrates the use of MLM for research studies of individual psychotherapy, group psychotherapy, and rolling group psychotherapy and provides SPSS syntax for all three cases. Psychotherapy research can be thought to involve multiple levels of analysis: patient, therapist, and group. Patients are nested in therapists and therapists are nested in treatments. When therapists each treat two or more patients, as is typical in studies of psychotherapy process and outcome, patients seeing the same therapist may be more similar to one another than they are to patients who see a different therapist. The magnitude of dependence in scores attributable to a nested factor is quantified by the intraclass correlation coefficient. There are some major advantages in using MLM over the traditional ANOVA method. Currently, most MLM programs including SPSS can handle dichotomous outcomes.
