Objective: The present study investigates skill transfer from Virtual Reality (VR) sports training to the real world, using the fast-paced sport of table tennis.
Background: A key assumption of VR training is that the learned skills and experiences transfer to the real world. Yet, in certain application areas, such as VR sports training, the research testing this assumption is sparse.
Design: Real-world table tennis performance was assessed using a mixed-model analysis of variance. The analysis comprised a between-subjects (VR training group vs control group) and a within-subjects (pre- and post-training) factor.
Method: Fifty-seven participants (23 females) were either assigned to a VR training group (n = 29) or no-training control group (n = 28). During VR training, participants were immersed in competitive table tennis matches against an artificial intelligence opponent. An expert table tennis coach evaluated participants on real-world table tennis playing before and after the training phase. Blinded regarding participant's group assignment, the expert assessed participants' backhand, forehand and serving on quantitative aspects (e.g. count of rallies without errors) and quality of skill aspects (e.g. technique and consistency).
Results: VR training significantly improved participants' real-world table tennis performance compared to a no-training control group in both quantitative (p < .001, Cohen's d = 1.08) and quality of skill assessments (p < .001, Cohen's d = 1.10).
Conclusions: This study adds to a sparse yet expanding literature, demonstrating real-world skill transfer from Virtual Reality in an athletic task.