You Like This Robot? I Don't! How Individual Differences Influence Perceptions of Robot Teammates in Virtual Reality

Abstract

As multimodal artificial intelligence (AI) systems increasingly appear in collaborative settings, understanding how individual differences shape perceptions of embodied agents is essential, particularly for agents capable of taking on multiple forms. This exploratory study investigates how personality traits and attitudes toward robots influence perceptions of re-embodied robot teammates in virtual reality (VR). Participants engaged in immersive scenarios with robots represented as machine-like, augmented, or human-like avatars. Results show that individual differences significantly moderated the robots' appearance suitability and teammate trustworthiness. Participants with more positive attitudes toward robots favoured robotic forms, while those with more negative attitudes showed an increased preference for human-like embodiments. Neuroticism was associated with increased trust in human-like avatars but reduced trust in robotic forms. These findings highlight the limitations of universal embodiment strategies for robot teammates and support the development of adaptive, user-aware design in multimodal AI systems. Further confirmatory studies are needed to validate these patterns across broader contexts and populations.