Year: 2026
Type: article
Abstract: Introduction Because negative treatment outcome expectations predict unfavorable treatment effects, violating these negative outcome expectations is one central goal of psychotherapy. One mediator of outcome expectations and therapy outcome is the therapeutic alliance. Methods This experimental online study investigated whether human vs. photorealistic avatar therapists can establish a therapeutic alliance through a first empathy video followed by an expectation violation video, challenging negative treatment expectations. Healthy participants ( N = 484) were assigned to one out of four experimental groups where either a human therapist or their photorealistic avatar expressed empathy (video 1) and gave expectation-violating information (video 2). Results In line with our hypothesis ( two human therapists > human , then avatar therapist > avatar , then human therapist > two avatar therapists) , outcome expectation improvements and alliance ratings were highest after the two videos for the sequence with two human therapists, and still considerably positive in the groups with the avatar in the first video. Discussion Contrary to our hypothesis, outcome expectations and alliance ratings declined after the second video for the group that first saw a human and then an avatar therapist: rather than potentiating the expectation violation after a human alliance, a negative contrast effect could have occurred when seeing an avatar after an initial human therapist, besides a modest, general effect of avatar-delivered content. When combining human and avatar therapist modalities in a single session, it seems important to consider the sequence of avatar and human therapist display: targeted expectation violation effects may be stronger when the human therapist follows the avatar.
Source: Frontiers in Psychology
Authors: Pauline Bauermann, Winfried Rief
Institution: Philipps University of Marburg
Language: English
Subfield: Clinical Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: gold
Estimated APC paid: $3,801
Funder: Deutsche Forschungsgemeinschaft
OpenAlex ID: W7223089191