The Role of Viewing Distance and Feedback on Affordance Judgments in Augmented Reality

The effectiveness of Augmented Reality (AR) for spatial applications is likely influenced by the extent to which users perceive that they can act on virtual objects as if they are real. While there are some promising initial results that action capabilities in AR are perceived similarly to the real world, AR differs from completely real-world viewing both in the visual information available within a restricted field of view (FOV) and in the ways that perceptual-motor feedback can be provided for the outcome of actions. We addressed both of these differences in a study of judgments of passing through an AR aperture defined by two moveable virtual walls presented via the Microsoft HoloLens. First, we investigated the possible effects of viewing distance on affordance judgments, predicting greater error in judgments at a close distance because of the inability to see the entirety of the aperture within the FOV. Second, we explored the effect of feedback on affordance judgments, predicting that verbal feedback about whether a judgment was correct would be sufficient to see improvements in judgments over time. In contrast to our predictions for distance, we found that judgments for passing through were closer to actual shoulder width when viewed at a distance near the aperture (0.85 m) compared to a farther viewing point (3.20 m). Verbal feedback reduced error over trials, but only at the farther distance, possibly because the judgments at the near distance were close to the ceiling of performance. These findings have implications for tasks that require accurate perception of action capabilities in AR, such as training procedures.