User Experience and Socioemotional Evaluation of Virtual Therapist Character Types in a Simulated Virtual Reality Exposure Therapy: Quantitative Study.

Abstract

Background: Virtual reality exposure therapy (VRET) has emerged as an effective treatment for individuals with specific phobias, particularly for those unable to access traditional exposure therapy. However, several aspects of VRET applications still require further investigation, including the role of embodied conversational agents as virtual therapists.

Objective: This study aims to investigate how the realism level and species of a virtual therapist influence user experience and perceived socioemotional relationship skills in a VRET intervention for dog phobia.

Methods: We conducted a quantitative study that examined user experience and perceived socioemotional relationship skills of virtual therapist character types, where participants were recruited from a convenience sample and were not required to have a dog phobia. In 4 simulated VRET interventions, participants interacted with 4 virtual therapists through a role-play dialogue (3.5 minutes per scenario), followed by an exposure scenario with a dog. The 4 virtual character types included a realistic adult woman, a stylized adult woman, a stylized cat, and a stylized robot. We collected quantitative data through standardized questionnaires assessing user experience and presence, cybersickness, trust, and willingness to share information, and an exploratory, theory-informed questionnaire assessing the virtual therapist's socioemotional relationship skills.

Results: A total of 40 participants completed the experiment. The realistic human scored significantly lower in hedonic quality than the stylized human (z=2.53; P=.01; P _adj=.02; rbb=-0.54), the robot (z=2.726; P=.006; P _adj=.01; rbb=-0.56), and the cat (z=3.772; P<.001; P _adj=.004; rbb=-0.81). Additionally, the cat scored higher in hedonic quality than the stylized human (z=2.528; P=.02; P _adj=.02; rbb=-0.52). While no significant differences were found among the virtual character types regarding trust or willingness to share information, the therapeutic relationship skills of the robot were rated significantly lower than those of the stylized human (z=-2.315; P=.02; P _adj=.02; rbb=-0.42). However, when asked to indicate a preferred virtual therapist, 30% (12/40) of the participants chose the robot, 27.5% (11/40) the stylized human, 25% (10/40) the cat, and 17.5% (7/40) the realistic human, making the latter the least favored preferred option in our study sample.

Conclusions: The realistic human was the least preferred virtual therapist, whereas the stylized human received more favorable evaluations. This finding suggests that increased visual realism does not necessarily result in more positive perceptions of virtual therapists and may, in some cases, lead to less favorable user responses. The cat's higher hedonic quality and the robot's popularity suggest that user perceptions may vary across different virtual therapist embodiments. These findings underscore the need for further research on virtual therapist design in VRET.