Effect of a Body Model on Performance in a Virtual Environment Search Task

Abstract

Abstract : The U.S. Army Research Institute is investigating requirements for using Virtual Environments (VE) in training dismounted soldiers. This experiment investigated full body representation (generic) versus a hand linked pointer on movement performance in an office building interior during a search task. The search task was used as a representative dismounted soldier activity in urban environments. The VE used a biocular Head Mounted Display (HMD) with head coupled and body referenced movement control. Sensors enabled participants to walk through the VE while performing the search task in six repeated trials. Movement time and number of collisions during discrete phases of the search task revealed no significant differences found between full body and pointer representations, although significant improvement was found over repeated trials. Field of view is discussed as a possible intervening aspect. A Simulator sickness questionnaire (SSQ) was administered before, during, immediately after the experiment, and after a recovery period. Significant changes in the SSQ were found over the course of the experiment, but were not related to the body representation condition. The results indicate a rapid onset of symptoms followed by some adaptation to the VE, and rapid recovery. The Immersive Tendencies Questionnaire administered pre-experiment, and the Presence Questionnaire administered post-experiment, were not significantly related to the body representation conditions.