Purchase this article with an account.
Amy A. Kalia, Paul R. Schrater, Gordon E. Legge; Comparing the reliability of vision and walking for target localization in a hallway. Journal of Vision 2009;9(8):1129. doi: 10.1167/9.8.1129.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
When navigating in an environment, humans can learn the locations of objects (e.g., rooms or landmarks) with two types of sensory cues: visual information obtained by viewing the objects, and information obtained by walking to the objects (proprioceptive and vestibular). We are interested in whether humans weight these two types of information according to their reliability when learning locations in a real hallway. Reliability is defined as the inverse variance (1/σ2) of the localization estimates. As a first step towards a cue conflict study, we compared the reliabilities of different levels of visual information (normal and blurry) and walking information in a spatial localization task. In the visual condition, participants viewed targets in a hallway with overhead lighting. Viewing was monocular with either normal or blurry (Snellen acuity of approximately 20/675) vision. Targets were high-intensity Light-Emitting Diodes located 5 to 11 meters away from the observer. After viewing a target, participants (N = 3) wore a blindfold and indicated the target's location on a tactile map of the hallway. In the walking condition, blindfolded participants (N = 3) walked until stopped by the experimenter (distances of 5 to 11 meters), and then indicated their location on the tactile map. In both conditions, map estimates were scaled to obtain equivalent distances in the real space. Results indicate that estimates made with blurry vision (σ2 = 1.08 m) were less reliable than with normal vision (σ2 = 0.16 m), but were more reliable than estimates made after walking to locations (σ2 = 1.79 m). These results predict that when both visual and walking information are available, observers will weight information from walking more heavily in the blurry viewing condition compared with the normal viewing condition. Preliminary results support this prediction.
This PDF is available to Subscribers Only