Sophia R. Baia¹ (), Michael K. McBeath^1,2; ¹Arizona State University, ²Max Planck Institute for Empirical Aesthetics

Introduction: Previous research found participants exhibit a systematic bias to overestimate the angle that their body is tilted, and that this bias increases under conditions where it is more costly or difficult to prevent falling. The current study investigates if this illusion is consistent with the “What” vs “Where/How” system by comparing body angle estimates of 45° to the maximum body angle at which individuals cannot prevent a fall by taking a step forward or backward. Method: 25 Participants leaned forward and backward to determine the maximum body-tilt angle before needing to take a step. They then had their body tilted in a body-sized gimbal gyroscope in which they estimated both when they felt tilted 45° and when they felt tilted at the angle needed to take a step. We also measured individual differences in balance ability as a potential mediating covariate. Results: We found that generally participants accurately estimated their body-tilt angle at which they would need to take a step (within 1° forward and 4° backward), but significantly underestimated when they were tilted 45° (by an average of 14°). We also found that individual balance metrics were not significant mediators of any body-tilt estimation effects. Discussion: Participants were accurate in estimating the angle of body-tilt needed to take a step, consistent with usage of the Where/How System (analogous to accurately estimating hill slant with a palm board). They also verbally underestimated their angle of 45° body-tilt, consistent with usage of the What System (analogous to verbally overestimating hill slant angle). This pattern of findings along with the lack of individual differences due to balance ability and the previous pattern of findings support that the body-tilt illusion is a functional distortion that likely helps people better keep their balance with little cost due to feeling more tilted than actual.