Motion blending in avatar-mediated communication does not necessarily increase interpersonal attraction

Avatar-mediated communication in virtual reality (VR) typically maps users’ movements onto their avatars, yet behavioral interventions can be introduced to modulate social perception. Here we tested whether pseudo-mimicry-like bodily similarity induced by motion blending, that is, a technique that generates the partner’s avatar by a weighted average of the participant’s and the partner’s movements, enhances interpersonal attraction. Twenty-four participants formed unfamiliar dyads and performed an idea-generation conversation task in an immersive VR room. From each participant’s viewpoint, the partner’s avatar reflected the participant’s own movements at 0%, 25%, 50%, or 100% blending ratios (within-subjects). After each conversation, participants rated the partner using a multidimensional interpersonal attraction questionnaire (social, physical, and task attraction). Trait empathy was assessed using the Interpersonal Reactivity Index (IRI). Friedman tests showed no significant main effect of blending ratio on any attraction dimension. In an exploratory analysis, Perspective Taking (IRI) showed a positive association with social attraction at the 50% blending condition at the uncorrected level, but this association did not remain significant after correction for multiple comparisons. Post-experimental debriefing further suggested that the manipulation was highly salient to most participants. These findings indicate that motion blending-based pseudo-mimicry-like bodily similarity did not produce a robust increase in interpersonal attraction in VR avatar communication under the present conditions, and highlight the need to consider detectability, temporal dynamics, and individual differences when designing behavioral interventions in social VR.