Effects of stress on perceptual learning in a virtual reality environment

Introduction: This study investigates two key questions in perceptual learning using a ten-day visual training protocol embedded in a first-person shooter (FPS) task within virtual reality (VR): (1) whether training improves the visual system’s ability to integrate orientation information across the visual field, and (2) whether aversive electrodermal stimulation and associated stress levels influence perceptual performance and/or learning. Method: 17 participants successfully completed an orientation-averaging task involving Gabor arrays of varying set-sizes (1, 2, 4, or 8 elements) under three shock conditions: no shock, performance-contingent shock, and random shock. Results: Training led to improvements in both accuracy and response times, while increasing set-size monotonically impaired performance. The interaction between training and set-size was weak, suggesting that training benefits likely emerge at a post-integration or decisional stage. Stress responses, indexed by the State-Trait Anxiety Inventory (STAI), confirmed elevated anxiety in both shock conditions compared to control. However, this increase in state anxiety did not reliably affect task performance or learning outcomes, nor did it modulate set-size effects. Notably, participants’ accuracy degraded when an on-screen health bar depicted lower “health” levels, regardless of actual shock delivery or contingency. Discussion: These findings indicate that visual feedback cues can shape engagement or motivation independently of experienced stress. More broadly, the results underscore the role of motivational and attentional mechanisms in immersive training environments and provide a framework for evaluating stress effects using subjective anxiety and objective psychophysical measures.