Catching virtual throws: An immersive virtual reality setup to evaluate human predictive skills

We present and validate a novel and portable IVR setup conceived for studying the predictive mechanisms associated to action observation. The setup implements an interactive throwing-catching task in which participants have to intercept balls thrown from a virtual character. To validate the setup, we performed a preliminary experiment in which participants had to intercept balls thrown by different throwers, under different ball/thrower visibility conditions. Non-expert adult participants were able to extract information from an observed throwing action to improve their interceptive performances. This ability was modulated by the throwing strategy (e.g. throwing from a fixed stance with respect to throwing with stepping corresponded to worse interceptive performances). These preliminary results validate our setup as a novel tool for exploring how humans access and make use of information from observed actions to optimize interpersonal interactions. Importantly, the proposed setup could be used as a tool for early diagnosis of pathologies in which predictive skills are progressively impaired.