Previous research has drawn attention to the fact that Virtual Reality (VR) technology is primarily designed around the interests and needs of male, western, educated, and wealthy individuals. In this paper, we look specifically at the avatars used in VR research. We analyzed 40 recent studies in which research participants embodied full avatars or were given virtual hands. While nearly every study matched the gender of the participant to the gender of the avatar, similar efforts were not made regarding skin color. We draw attention to two serious problems in VR research: the explicit and implicit exclusion of participants on the basis of skin color through avatar design. Explicit exclusion, in which only participants whose skin tones matched the light-skinned avatar were recruited occurred in 5% of the studies analyzed. Implicit exclusion occurred when participants embodied an avatar whose skin tone did not match their own. Only 10 (17.5%) of the studies analyzed provided avatar personalization which included different skin tone options that participants could select. Some of the only studies (five out of 40) which utilized Black or dark-skinned avatars were doing so in the context of studying racial bias with all white participants. We argue that using light-skinned default avatars is not acceptable from a research ethics standpoint and that doing so serves to further maintain and normalize systemic racism. We provide recommendations for researchers with the goal of both mitigating and bringing awareness to current issues in VR research.

Mental rotation, a common measure of spatial ability, has traditionally been assessed through paper-based instruments like the Mental Rotation Test (MRT) or the Purdue Spatial Visualization Test: Rotations (PSVT:R). The fact that these instruments present 3D shapes in a 2D format devoid of natural cues like shading and perspective likely limits their ability to accurately assess the fundamental skill of mentally rotating 3D shapes. In this paper, we describe the Virtual Reality Mental Rotation Assessment (VRMRA), a virtual reality-based mental rotation assessment derived from the Revised PSVT:R and MRT. The VRMRA reimagines traditional mental rotation assessments in a room-scale virtual environment and uses hand-tracking and elements of gamification in attempts to create an intuitive, engaging experience for test-takers. To validate the instrument, we compared response patterns in the VRMRA with patterns observed on the MRT and Revised PSVT:R. For the PSVT:R-type questions, items requiring a rotation around two axes were significantly harder than items requiring rotations around a single axis in the VRMRA, which is not the case in the Revised PSVT:R. For the MRT-type questions in the VRMRA, a moderate negative correlation was found between the degree of rotation in the X direction and item difficulty. While the problem of occlusion was reduced, features of the shapes and distractors accounted for 50.6% of the variance in item difficulty. Results suggest that the VRMRA is likely a more accurate tool to assess mental rotation ability in comparison to traditional instruments which present the stimuli through 2D media. Our findings also point to potential problems with the fundamental designs of the Revised PSVT:R and MRT question formats.