Social virtual reality (VR) applications enable users to experience highly immersive mediated interactions and social experiences. Especially embodied nonverbal communication, like mediated social touch, is mostly unique to social VR and potentially a major factor in enabling fulfilling emotional experiences at a distance. In our within-subjects study ($N = 28$), we compared feedback from wearable vibrotactile haptic devices with two established visual presentation forms of social touch: clipping and pseudo-haptics. Our findings suggest that the addition of haptic feedback can elicit significantly higher emotional arousal and result in a stronger illusion of touch than visual presentations alone. In contrast to existing literature, there was no evidence of a difference between the immersive and emotional experience of clipping and pseudo-haptics in our sample. In addition, we also investigated whether simply wearing haptic feedback devices during a social touch interaction, i.e., without active haptic feedback, affects the illusion of touch and user experience. However, there was no evidence of a significant effect in the sample studied. A free copy of this paper, including the appendix, is available at https://doi.org/10.17605/OSF.IO/4FYKQ.

Figure 1: Player avatars search the virtual escape room for clues.

The high degree of sensory immersion is a distinctive feature of head-mounted virtual reality (VR) systems. While the visual detachment from the real world enables unique immersive experiences, users risk collisions due to their inability to perceive physical obstacles in their environment. Even the mere anticipation of a collision can adversely affect the overall experience and erode user confidence in the VR system. However, there are currently no valid tools for assessing collision anxiety. We present the iterative development and validation of the Collision Anxiety Questionnaire (CAQ), involving an exploratory and a confirmatory factor analysis with a total of 159 participants. The results provide evidence for both discriminant and convergent validity and a good model fit for the final CAQ with three subscales: general collision anxiety, orientation, and interpersonal collision anxiety. By utilizing the CAQ, researchers can examine potential confounding effects of collision anxiety and evaluate methods for its mitigation.