Abstract
The ability to recognize and use object landmarks is crucial for efficient wayfinding. However, the roles of different types of landmarks in wayfinding remain to be clarified. In the current study, we examined how global landmarks and local landmarks were utilized differently in a virtual maze environment. There were three experimental conditions: a global-landmark condition with tall architectures that surrounded the virtual maze and were visible from almost everywhere, a local-landmark condition with cartoon pictures of common objects that were posted on the walls along the paths inside the maze and were only visible from certain locations, and a control condition that had no object landmarks. The participants learned the layout of the maze and the target positions through a series of trials. Their learning performance was assessed by traveling distance, time, and ratio of successfully found targets within time limits. The results of Experiment 1 demonstrated that while participants learned to navigate most efficiently in the local-landmark condition, their performance did not differ between global-landmark and control conditions. In Experiment 2 we increased the number of target locations for navigation in each trial, and the results showed that participants could benefit from both global- and local- landmarks as opposed to the control condition, though they still performed better in the local-landmark condition than in the global-landmark condition. Taken together, local objects seem to be easier to use as reference points than global objects. Global-object landmarks can still be used to assist wayfinding, but they may become useful only when there is high demand on memory of target locations.
This research project was funded by National Science Council, Taiwan (Grant ID#: NSC96-2413-H008-003-MY3 and NSC95-2627-E009-001).