Psychotherapy.

The present study investigates the affectual interaction of three highly competent
therapists using different theoretical approaches with a shared male client and female
client in a professionally recorded video series. The interactions of clients and therapists
in a total sample of six psychotherapy sessions were coded using the twenty code version
of Gottman, Woodin, and Coan’s (1998) Specific Affect Coding System. Coded data
were analyzed using Kruskal-Wallis tests which found no significant differences between
the mean ranks of therapists’ codes between therapists. Significant effects among some
affective behaviors were noted in therapists’ codes when compared by client. Coded data
were also used to create mathematical models using ordinary differential equations for
each of the six sessions. Kruskal-Wallis tests did not reveal significant effects in the
mean ranks of the parameters of the mathematical models, and visual similarities and
differences of these models were discussed. Additional analyses were conducted to examine clients’ affective behaviors as well and significant effects were revealed in the
Kruskal-Wallis tests amongst many coded behaviors. The results support the presence of
common factors and similarities in the therapeutic relationship among different
theoretical approaches. The findings also add to the growing body of literature dedicated
to the use of observational coding and dynamic nonlinear modeling in psychotherapy
research. The implications for psychotherapy practice, education, and research are
discussed.

The present study applies Driver and Gottman’s (2004a) Turning System to
observations of the therapeutic relationship in a quasi-experimental between and within
groups design. A sample of 63 full counseling sessions (21 first sessions with clients who
return for four sessions, 21 first sessions for clients who terminate therapy prior to four
sessions, and 21 fourth sessions) were collected from a university counseling center in
South Florida. Clients and clinicians also completed self report evaluations of the
therapeutic relationship that were also included in this study (Working Alliance Inventory
– Short Form, and the Real Relationship Inventory). A series of multivariate analysis of
the variance (MANOVA) tests were performed to assess for significant differences in
Turning System behavior between return and dropout groups in the first sessions of
therapy, as well as for significant differences between return groups’ first and fourth sessions. Correlation analyses were run for client and therapist self report data and
Turning System codes.
Overall, the Turning System codes did not predict attrition from therapy;
however, significant effects were found for specific behaviors in the return versus
dropout comparisons, including client’s uses of negative bids, high level questions, and
preoccupied away responses. The Turning System also did not predict clear differences
between behaviors in ongoing therapy, though significant effects were again found for
individual behaviors enacted by both clients and therapists. Significant negative
correlations were also found for return group clients between specific behaviors and
ratings of the therapeutic alliance, such as high level questions and preoccupied away
responses. Individual behaviors in the therapist return group, such as high level questions,
negative bids, and interruptions, correlated negatively with ratings of the therapeutic
alliance. The results of this study are presented in an effort to synthesize the data into a
narrative for developing effective therapeutic relationships and guidance for future
research.