Telepresence and Performance in an Immersive Virtual Environment and Sporting Task

Abstract

The purpose of this research was to assess the relationship between telepresence and performance in a synthetic environment. Telepresence is believed to be a mental construct and to enhance task performance in teleoperations and virtual environments. Consequently, it has been identified as a design ideal for synthetic environments. However, there is a limited understanding of telepresence and its relation to task performance. This research involved examination of a range of synthetic environment design features (e.g., viewpoint and auditory cue type) that were suspected to influence telepresence and compared differences in telepresence and task performance caused by manipulations of these factors and task difficulty.

A simulated basketball free-throw task was used in which subjects controlled the motions of a virtual basketball player. In addition to the basketball task performance (baskets/goals), subjects were required to report camera flashes in the virtual environment (stadium) and to simultaneously detect strobe light flashes in a real research laboratory. These tasks were designed as secondary-monitoring tasks and were intended to assess subject attention allocation to the virtual and real environments as an indicator of telepresence. Each subject was exposed to a single viewpoint condition including either an egocentric view, an exocentric view from behind the player, an exocentric view from the sideline of the court, or a selectable viewpoint. They were also exposed to four virtual sound conditions including task-relevant sounds, task-irrelevant sounds, a combination of sounds and no sound, as well as two visual display fidelity conditions including a low fidelity stadium composed of rendered walls surrounding the basketball court and a high fidelity stadium that displayed a texture of a crowd watching the game. Finally, the subjects experienced two task difficulty conditions including 2-point and 3-point shots. The order of presentation of the sound, fidelity and difficulty conditions was randomized. Subjective ratings and the objective, attention-based measure of telepresence were recorded during the experiment, along with task performance and workload.

The results of the study provided evidence that the features of a simplistic synthetic environment, which include immersiveness (viewpoint) and auditory cue type significantly influence the sense of subjective telepresence. However, the objective, attention-based measure of telepresence did not prove to be sensitive to the experimental manipulations. Virtual task performance was significantly affected by task difficultly. This study also revealed significant effects of viewpoint and audio cue type on subjective workload. An analysis of trends on changes in telepresence and performance across settings of various virtual reality features provided compelling evidence that telepresence is a predictable experience unique from performance. However, the results of this analysis cannot be considered conclusive in terms of describing a causal relationship between telepresence and performance, in part, because of mixed findings across predictors. Beyond the relationship of telepresence and performance, this study provided further evidence of significant positive relations between telepresence and workload. There were no significant interaction effects among the virtual reality factors mentioned above in terms of performance, telepresence and workload. Although counter to expectation, and some previous research hypotheses on cross-modality interactions in virtual reality experiences, this finding was consistent with the findings of other prior empirical work. The lack of interaction effects on the response measures suggests that virtual reality design using multiple sensory channels could be relatively simple. However, this study did not examine all types of sensory cues including tactile and haptic. Furthermore, it is possible that more complex virtual environments for real-world applications might cause telepresence experiences more sensitive to, for example, cue conflicts.