Abstract
Perceptual matches between vertical extents (poles) and egocentric extents (between the observer and the pole) show that observers set themselves much too far from the pole, consistent with an under-perception of egocentric distance (Li, Phillips & Durgin, 2010). These matches can be modeled by assuming that angular gaze declination is exaggerated with a pitch-angle gain of 1.5. Matches to horizontal frontal extents suggest a lesser yaw-angle gain of about 1.2 (Li et al., in press). We tested whether angular biases in space perception were affected by observer orientation relative to vertical. Observers (96) were tested in four, between-subject conditions: (1) Walking, (2) Sitting upright on a cart, (3) Lying sideways on a cart (tilted 90° from vertical), and (4) Lying at an oblique angle (54° from vertical) on a cart. Each observer made three judgments: one egocentric matching judgment to a 10 m vertical pole (half started near; half far and adjusted themselves to be in the apparent match location), one 45° gaze elevation judgment to a 35 m tower (half started near; half far and adjusted themselves to be at an apparent 45° to the top of the tower), and one verbal height estimate of the 35 m tower. Upright observers and tilted observers showed similarly biased matches between egocentric distance and object height and consistently biased apparent 45° gaze setting, consistent with the model proposed by Li et al. (2010). This suggests that exaggerations of gaze elevation and declination are referenced to the world rather than the body. However, tilted observers gave reliably lower verbal estimates of tower height (geometric mean: 28 m), than did upright observers (45 m). Although eye-height was similar across conditions, it may have been underestimated in the tilted conditions -- which should reduce height estimates proportionally but not affect matching.
Meeting abstract presented at VSS 2013