Bias and accuracy in self-assessments of psychotherapeutic skills: A multi-perspective secondary analysis

Objective: Self-assessment of psychotherapeutic skills is assumed to be prone to bias. The present study aimed to investigate psychology students' ability to assess their own psychotherapeutic skills, and to compare it with the assessment from independent raters and supervisors. Method: Data from 73 psychology students were analyzed from a randomized controlled trial investigating the effect of clinical supervision on skill development. Students' basic communication skills, therapeutic alliance, and empathy were assessed by independent raters and students themselves. Supervisors assessed students' basic communication skills and empathy after supervision. A truth and bias model was applied to examine bias and accuracy in students' self-assessment. Results: Across outcomes, the results did not indicate a consistent pattern of overestimation (self-assessment bias). Students only overestimated their therapeutic alliance in comparison with independent observers, and underestimated their empathy in comparison with supervisors. Overall, students' self-assessments did not yield significant accuracy in relation to independent observer or supervisor ratings. Conclusions: Self-assessments of students in the early stages of their training may lack accuracy compared with external assessments, while the role of systematic bias needs further investigation. Future research should explore methods, such as feedback, to improve the precision of self-assessment.