A randomized controlled trial of a feedback method for improving empathic accuracy in psychotherapy
Article first published online: 14 MAR 2011
©2010 The British Psychological Society
Psychology and Psychotherapy: Theory, Research and Practice
Volume 84, Issue 2, pages 113–127, June 2011
How to Cite
Sripada, B. N., Henry, D. B., Jobe, T. H., Winer, J. A., Schoeny, M. E. and Gibbons, R. D. (2011), A randomized controlled trial of a feedback method for improving empathic accuracy in psychotherapy. Psychology and Psychotherapy: Theo, Res, Pra, 84: 113–127. doi: 10.1348/147608310X495110
- Issue published online: 13 MAY 2011
- Article first published online: 14 MAR 2011
- Received 21 November 2008; revised version received 5 February 2010
Objectives. To develop and evaluate a feedback method for reducing empathic errors in psychotherapy.
Design. Randomized controlled trial conducted in a university-affiliated out-patient psychiatric clinic.
Methods. Sixteen non-psychotic patients treated for Axis I disorders by 12 psychiatry residents were randomly assigned to intervention and control conditions. In both conditions, at the end of each session, patients rated their own functioning on the Global Assessment of Functioning scale, and therapists predicted patients' ratings. Patients predicted their therapist's accuracy and therapists rated their confidence in their own predictions. In the intervention condition, therapists and patients reviewed their respective ratings from the previous session together. In the control condition, ratings were given directly to the investigator without being reviewed by either patients or therapists.
Results. Therapists in the intervention condition showed greater overall accuracy than controls as well as evidence of increasing empathy later in therapy on the Barrett-Lennard empathy subscale. Patients in the control group perceived their therapists as significantly more or less accurate than was warranted according to the accuracy measure (over-/under-idealization). Therapists in the control group were more likely than those in the intervention group to overestimate their own accuracy (overconfidence). Affective responses to the instrument were positive overall and did not differ by condition.
Conclusion. An intervention such as the one tested in this study may be a practical and useful method for improving accuracy of understanding in a variety of training and clinical settings.