Immersive VR for the Assessment of Spatial Skills in Adolescents: Performance, Gender Effects, and Links to Computational Thinking

Spatial skills, in particular mental rotation, are increasingly linked to STEM achievements and to computational thinking (CT) abilities. Traditional 2D spatial assessments face validity and equity issues from visual ambiguity, occlusion, and missing depth cues, requiring cognitively demanding 2D-to-3D interpretation. These artifacts might contribute to gender disparities, giving males an advantage that may reflect test bias rather than true ability differences. Immersive Virtual Reality (VR) seems to mitigate these issues by providing stereoscopic depth, multi-perspective viewing, realistic lighting, and embodied interaction. We compared spatial performance in VR and traditional 2D assessments using the Virtual Reality Mental Rotation Assessment (VRMRA) in a within-subjects study of 48 adolescents (ages 12-16). VR improved performance, yielding higher accuracy in Mental Rotation Test (MRT)-style tasks (+1.3 items, p < .001), while PSVT:R accuracy did not differ significantly between 2D and VR. We observed that gender effects are task-specific: VR improved the performance of females most in MRT-style tasks, reversing the 2D male advantage, whereas males gained more in PSVT:R-style tasks. Spatial scores correlated positively with CT across both 2D and VR assessments, indicating a link to computational problem-solving. These results suggest that the assessment of spatial skills in VR produces distinct performance patterns with respect to traditional 2D assessment, highlighting the importance of immersive VR for the development of improved tools and its potential implications for inclusive STEM education and talent identification.