Modeling in psychotherapy training: A randomized controlled proof-of-concept trial.

Objective: Despite the importance of modeling for learning, it has only so far been investigated in a few controlled studies in psychotherapy training. We, therefore, investigated, across several study outcomes, whether modeling outperformed a control group.

Method: Sixty-nine trainees (i.e., psychology students; mean age 25.58 years, 81.2% women and 81.2% Caucasian) were randomly assigned to the intervention group (IG, manual reading plus modeling) or control group (CG, manual reading). After manual reading, IG participants watched a video of a skillfully conducted therapy session, whereas CG participants watched an unspecific tutorial (e.g., solving a Rubik's Cube). Trainees then demonstrated cognitive behavioral interventions in videotaped role-plays with (SPs). Psychotherapeutic competences and counseling skills were rated by two independent raters, on the Cognitive Therapy Scale (CTS) and the Clinical Communication Skills Scale-Short Form (CCSS-S). In addition, raters, trainees, and SPs assessed empathy (Empathy Scale [ES]) and alliance (Helping Alliance Questionnaire [HAQ]).

Results: Multilevel modeling revealed Significant Time × Group Effects, that is, psychotherapy competences (CTS, β = .26, p = .026) and counseling skills (CCSS-S, β = .31, p = .004) to increase more pre-post in the IG than in the CG (rater perspective). The same applied to the alliance (HAQ) from the external rater and SP perspectives. Trainees evaluated the alliance as improving in both groups from pre to post training (main effect).

Conclusions: The results provide initial support for modeling as important for fostering psychotherapy skills. Future studies should focus on more complex training and real patient encounters. (PsycInfo Database Record (c) 2023 APA, all rights reserved).