Abstract
Past research has shown that people can judge their instantaneous direction of travel (heading) along linear trajectories within 1 (Warren & Hannon, 1988). However, viewing a curvilinear trajectory introduces biases in heading judgments of up to 5 in the direction of rotation (Stone & Perrone, 1997). In the current experiment, subjects actively steered a path consisting of straight, low curvature (6 /s), and high curvature (12 /s) segments. The subjects' task was to stay over the centerline of a path 2.2 eye-heights wide. They moved at a constant speed of 2.2 eye-heights/second, and their steering direction was perturbed by 3 levels of a sinusoidal wind. While steering these paths, subjects were occasionally asked to judge their heading. Heading judgments along straight paths were within 1 of the veridical. Although subjects were able to judge their heading fairly accurately while driving the straight path segments, the curved path segments introduced large biases of 4 and 5 in the direction of the curve for the low and high curvature segments, respectively. Additionally, heading judgment precision dropped by ∼30% when subjects were driving the curved segments. Despite the heading biases on curved path segments, steering performance (assessed by RMS error scores) was only slightly affected. For example, comparison between straight and low curvature segments shows only a minor decrement in steering performance (RMS error increases from 0.16 to 0.19 eye-heights) but a large decrement in heading perception (∼4 bias in the low curve condition).