Datensatz

Zusammenfassung

Planning a route with a destination not visible from the starting point is a behaviour only achievable with cognitive map like representations of space. Although important for defining cognitive maps, the mechanisms underlying route planning are widely unknown. This study attempts to investigate human route planning mechanisms using Virtual Reality techniques. Subjects navigated through a virtual environment that consisted of twelve places interconnected by roads. Six places were arranged in a hexagonal ring, the other places could be reached by dead-end roads starting from the corners of the hexagonal ring. While each place was uniquely specified by its associated landmarks, the places within the ring could be grouped into three different semantic clusters. A cluster consisted of neighbouring places with semantically related landmarks.
In order to investigate how many steps are planned in advance (depth of planning), subjects were asked to alternately navigate prime and probe routes. Prime routes were short routes that covered part of the subsequent longer probe routes. If a prime influences the planning and navigation of a subsequent probe route, the route must have been planned in advance at least up to that point where both routes overlap.
While the prime did not influence navigation of the probe route, an ANOVA revealed a main effect for semantic clusters (F=6.387; p=0.021). In 58 of the trials subjects chose the route connecting places within a semantic cluster rather than leaving the cluster (both possible routes were of equal length).
Although we could not prove that route planning is primed by prior navigation in this experimental setup, we demonstrated semantic priming in route choice experiments in a real-time navigation task. This effect might be considered a version of the Initial Segment Strategy known in route selection. Here subjects tend to choose routes that allow them to maintain the initially chosen direction as long as possible.