The effect of a virtual reality environment on gaze behaviour and motor skill learning

ObjectiveVirtual reality (VR) systems hold significant potential for training skilled behaviours and are currently receiving intense interest in the sporting domain. They offer both practical and pedagogical benefits, but there are concerns about the effect that perceptual deficiencies in VR systems (e.g. reduced haptic information, and stereoscopic display distortions) may have on learning and performance. ‘Specificity of learning’ theories suggest that VR could be ineffective (or even detrimental) if important differences (e.g. perceptual deficiencies) exist between practice and real task performance conditions. Nevertheless, ‘structural learning’ theories suggest VR could be a useful training tool, despite these deficiencies, because a trainee can still learn the underlying structure of the behaviour. We explored these theoretical predictions using golf putting as an exemplar skill.MethodIn Experiment 1 we used a repeated measures design to assess putting accuracy (radial error) and quiet eye duration of expert golfers (n = 18) on real putts before and after 40 VR ‘warm up’ putts. In Experiment 2, novice golfers (n = 40) were assigned to either VR or real-world putting training. Putting accuracy and quiet eye durations were then assessed on a real-world retention test.ResultsBoth visual guidance (quiet eye) and putting accuracy were disrupted temporarily when moving from VR to real putting (Experiment 1). However, real-world and VR practice produced comparable improvements in putting accuracy in novice golfers (Experiment 2).ConclusionOverall, the results suggest that: (i) underlying skill structures can be learned in VR and transferred to the real-world; (ii) perceptual deficiencies will place limits on the use of VR. These findings demonstrate the challenges and opportunities for VR as a training tool, and emphasise the need to empirically test the costs and benefits of specific systems before deploying VR training.