Abstract
Purpose: Goal-directed walking requires an estimation of the goal's location in space relative to one's body. Information for this estimation could come from various sources, e.g. visual, efferent, vestibular, and somatosensory signals. In this study we determined whether there is a gender difference in the effect of visual and non-visual information on action-task performance—walking to a briefly exposed target. Methods: Subjects were 10 men and 10 women. An immersive virtual environment was used to present a scene of a room with a ball in a doorway. The ball was positioned at −3 , 0 , or 3 relative to the subject's midline, and the doorway was either centered on the subject's midline or offset 5. The scene was displayed for 1 sec, replaced with a uniform field, and the subjects were to walk to the ball (4 m). In Exp. 2, subjects began walking at display onset. In Exp. 3, the positions of the subject's head and/or eyes were rotated relative to straight-ahead as they viewed the scene. Results: Doorway offsets affected the women's paths significantly more than the men's, t(18)=−2.3, p<.05. Nine of 10 women had errors of 2.8–22.7. The majority of the men (7/10) had errors of 0.1 – 1.4. With rotated head and/or eyes (Exp. 3), the women had significantly larger path errors than the men. With visual feedback (Exp. 2), the magnitude of the women's errors decreased to the range of the men's. Conclusions: Women are more affected than are men by the context of a target and by the positions of their head and eyes in determining the spatial position of a target to self.