What do "interpersonally sensitive" supervisors do and how do supervisees experience a relational approach to supervision?

OBJECTIVE: In two investigations, we identified explicitly relational supervision strategies and examined whether use of these strategies was associated with perceptions of the supervisory alliance and evaluations of the supervisor. METHOD: First, ratings by nine supervision researchers identified five clearly relational in-session strategies (focus on countertransference, exploration of feelings, attend to parallel process, focus on the therapeutic process, focus on the supervisory alliance) in the Critical Events model of supervision. Based on these expert ratings, we created and assessed the Relational Behavior Scale (RBS). RESULTS: Analyses with two samples of supervisees at all levels of training supported the measure's reliability and factorial validity. The RBS's validity was further indicated by its unique association with the "interpersonally sensitive" style of supervision. In both studies, supervisees perceived more frequent use of relational behavior on the part of psychoanalytic/psychodynamic/humanistic supervisors than cognitive-behavioral supervisors. Moreover, as hypothesized, supervisors' use of relational behavior in a specific session mediated the association between trainees' alliance perceptions and evaluations of their supervisors in that session. CONCLUSION: The identification of specific in-session supervision behaviors that explain one way in which a strong alliance contributes to trainees' positive experiences of their supervisors has implications for supervision theory, research, and practice.

Negotiating therapeutic alliances with a family at impasse.

To bridge the science-practice gap, the APA Presidential Task Force endorsed the publication of evidence-based case studies, but to date, there have been few such investigations of conjoint family therapy. To fill this gap, we studied a successful case of treatment-as-usual in a community agency. Owing to the complexity of the working alliance in conjoint therapy, we examined how an experienced family therapist managed to develop and sustain multiple alliances over time with an estranged couple in crisis. The outcome data showed clinically meaningful changes as well as high satisfaction levels and notable declines in the target complaint discomfort levels of all family members. Alliance indicators showed that the therapist worked diligently over time to connect emotionally with each family member and to foster and maintain safety. Session impact scores showed consistently deep sessions but more variability in smoothness. By working toward the only shared treatment goal-to repair each parent's individual relationship with their very angry daughter-the therapist was able to reduce the effect of the marital estrangement on the child. At the end of the 10 contracted family sessions, the parents agreed to begin working on their relationship in couples therapy, which led shortly thereafter to a reconciliation.

Exploring actor-partner interdependence in family therapy: whose view (parent or adolescent) best predicts treatment progress?

Predictions of family therapy outcome consistently vary depending on which client rates the alliance. We used the actor-partner interdependence model (Kenny, Kashy, & Cook, 2006) to test the interdependence of parents' and adolescents' ratings of alliance, session depth/value, and improvement-so-far after Sessions 3, 6, and 9. Initial analyses found trivial between-therapists variance; therefore, a 3-level hierarchical model partitioned the variance in these variables into between families, between family members, and between session components. For alliance and session depth, results showed a significant parent actor effect and a significant adolescent partner effect. Specifically, when parents saw a stronger alliance, they also saw the session as more valuable, but when adolescents saw a stronger alliance, their parents saw the session as less valuable. Both the parents' and the adolescents' improvement scores showed significant linear growth over time, and adolescents' alliance ratings were positively associated with their own and their parents' views of therapeutic progress.