Cassidy Brown¹, Star Niu¹, Frank Durgin¹; ¹Swarthmore College

Durgin and Gigone (2007), employed a head-mounted-display (HMD) on mobile observers to show that the perceived speed of optic flow was reduced during walking (relative to the same optic flow presented while standing). Moreover, perceptual discrimination of optic flow speeds near (and above) walking speed was enhanced while walking. These findings are consistent with sensory-prediction models of efficient coding, but were based on small numbers of participants (often 10 per condition). The present study sought to replicate some of these prior observations using more rigorous methods: a larger set of participants, more trials, and a wider field-of-view HMD. In an initial set of 22 participants, we observed strong mean reductions in perceived visual speed while walking such that 2.0 m/s during walking was matched to 1.55 m/s [95% CI: 1.45, 1.65], while standing. Although there was fairly high variance in this value, the effect size was large (D = 1.96). The mean difference in Weber fractions between discriminations made while standing (W = 0.073) and those made while walking (W = 0.055) at the mean walking speed (1.39 m/s) were in the predicted direction (p < .05), but had a smaller effect size (D = 0.40) in this fairly small sample. However, an unexpected effect of testing order suggested that the difference only emerges in the second block tested (presumably because of familiarity with the task), where Weber fractions for the Walking condition (W = 0.043) were significantly less than those in the second-block Standing condition (W = 0.074), t(20) = 2.87, p = .009 (D = 1.04). This replicates the prior observation that perceptual precision is enhanced. This occurs in naturalistic conditions where sensory information concerning expected flow can be derived from motor and vestibular signals.

Acknowledgements: Swarthmore College