Abstract
People are able to learn the spatial layout of an environment through navigation, but judgments about the relative position of locations often show large errors even after experience. Can viewing a top-down map of the structure of an environment help to learn a spatial layout more accurately? In a previous study, we found that previewing a top-down map before exploring a virtual environment produced no improvement in judgments of relative directions. In this study, we tested the effect of previewing a top-down map using a cue conflict paradigm. Before navigating through a virtual city block, subjects previewed a top-down map that was stretched along one of the main axes of the environment. If the distorted map were used to organize information learned during navigation, spatial judgments would be expected to show systematic biases. An experimental session consisted of three repetitions of map viewing, exploration, and testing. During exploration, subjects navigated in the virtual environment to find 8 targets. Spatial knowledge was tested by asking subjects to point toward each target from a central location, either with the environment visible or hidden. Pointing responses were used to find a best-fitting stretching transformation that represents the perceived distortion of the environment. More than half of subjects showed evidence of a systematic distortion of perceived spatial layout, but the distortions were idiosyncratic and unrelated to the distortion applied to the top-down map that was viewed. Subjects performed a second session in which procedure was repeated with accurate top-down maps of the same environment, and responses showed idiosyncratic biases with similar direction and magnitude. Our results provide no evidence that the map information was integrated with information acquired during navigation. The lack of integration may be due to difficulty translating between egocentric and map-based coordinates.