Abstract
Virtual environments (VEs) are computer-simulations of real or fictional environments with which users can interact. Potential applications of VEs include training, visualization, entertainment, design, rehabilitation, education, and research. For many applications, the utility of VEs increases if a person's actions and spatial perception in the VE are similar to those in the real world. Prior research, however, suggests that judgments of the distance between an observer and an external object, egocentric distance judgments , are systematically compressed in the VE compared to the real world. Ideally, we would like to understand the cause of this distance bias and correct for it. While many possible causes have been investigated, the reasons for the spatial bias are not yet fully known. One approach, taken in the present research, is to provide feedback to users in order to allow them to improve the accuracy of their behavior through a process called adaptation . This approach involves an empirical investigation that gives insight into the generalizability of adaptation and the perceptual processes involved in distance perception and adaptation. This insight is valuable from both a practical and a theoretical perspective.
In order to determine the practicality of providing feedback to the user within a VE and to gain insight into the process of spatial perception, we investigate three aspects of the generalizability of the effect of feedback. First, we examine whether the effect of perceptual-motor, unconscious coupling between perception and action, and cognitive corrective, conscious awareness of accuracy, aspects of feedback generalizes to both perceptual-motor and cognitive measures of perceived egocentric distance in an head-mounted display (HMD). We find that each of the three methods of feedback cause both egocentric distance judgments to be improved within an HMD. Second, we explore whether, with the same three methods of feedback, the effect transfers from the HMD to the real world. We find that real world egocentric distance judgments performed after experiencing the same visually matched HMD feedback are not altered. Third, in both an HMD and a treadmill-VE, we ask whether decoupling the speed of the visual flow from the walking speed causes real world egocentric measures of distance perception to be altered. We find that this visual decoupling manipulation causes a real world perceptual-motor egocentric distance task to be altered while a real world cognitive egocentric distance task is unchanged.
This research demonstrates that VE feedback can be used to improve egocentric distance judgments within VEs for those VE applications that require accurate spatial performance within action space. This research also determines that VE feedback affects subsequent spatial performance within the VE differently than it affects real world spatial performance. In addition, by investigating the generalizability of the effect of perceptual-motor and cognitive VE feedback within action space (2-25 meters) on perceptual-motor and cognitive responses in both the real and virtual world we are better able to infer the cause of the observed adaptation effects. At least four possible explanations could account for some, but not all, of the observed adaptation effects. First, perceptual-motor adaptation, a change in the coupling between the perception of the space and motor information, could account for some of the observed effects. Another cause might be a change in the perception of the space used to inform the response measures. Third, a higher-level cognitive decision rule may be selectively applied to some of the responses. Finally, a re-weighting of cues used for distance perception may occur due to VE feedback. We suggest that a higher-level cognitive decision rule or a re-weighting of distance cues, perhaps in combination with perceptual-motor adaptation is likely the cause of the observed adaptation effects within this dissertation.
