부산대학교 도서관

Psychological science increasingly employs virtual reality (VR) as a therapeutic and experimental tool. When represented by an avatar, a user's self-perception is altered through a full-body perceptual illusion known as virtual embodiment. However, inconsistent measurement of this construct derives uncertainties about the efficacy of VR-driven behavioural, cognitive, and social-psychological changes. Accordingly, we sought to provide psychological scientists a valid and reliable measure to assess the virtual embodiment construct during experimental and therapeutic simulations. In a repeated-measures experimental paradigm, participants (N = 86) completed a series of visuo-motor tasks in-vivo and in VR. Our study had three objectives: 1) compare visuo-motor performance across in-vivo and VR tasks; 2) assess the psychometric properties of theoretically foundational body ownership, agency, and self-location dimensions from a virtual embodiment questionnaire widely used in the literature; 3) analyse the simplified 11-item questionnaire's underlying structure using principal component analysis (PCA). We found that VR embodiment does not directly align with physical embodiment when considering proprioception and kinaesthesia, and there is limited support for convergent and discriminant validity of body ownership, agency, and self-location as independent dimensions of the embodiment construct. However, there is strong support for criterion validity when these dimensions comprise the overall measurement of embodiment. The PCA revealed a straightforward and theoretically interpretable underlying structure with three components: embodiment, disembodiment, and agency. This study offers psychometric support for the exploration of the simplified 11-item questionnaire as a theoretically grounded measure for behavioural scientists who seek to investigate virtual embodiment as a psychological construct.