Active route learning in virtual environments: disentangling movement control from intention, instruction specificity, and navigation control

This computation was chosen for continuity reasons throughout this research (i.e., the tour integration task used in Experiments 2–3). However, it does not exclude the possibility that a participant can first draw one wrong turn followed by correct turns, which results in a potential underestimation of this participant’s performance. We considered several alternative computation approaches, for example, to count whether two subsequent turns were correctly indicated (e.g., the original route included first a left turn, followed by a right turn. A correct indication of those turns would account one point). However, the overall quality of the sketched paths was rather low, so that it was impossible to determine which drawn turn corresponded to which turn of the original route. Due to the difficulties to establish a meaningful objective scoring criterion, we decided to rate subjectively how similar the sketched path was as compared to the original route (ranging from 1 = no similarity to 5 = absolute congruence; M = 2.23, SD = .86). The same ANOVA as reported in the results section showed no significant effects, all Fs < 1.18, ns. We conclude that participants of all experimental conditions were unable to deduce the cross-shape of the route from their egocentric encounter of the environment. In other words, they did not develop a cognitive map of the environment.

There was an interaction of intention and landmark relevance, F(1,78) = 4.49, p < 0.05, η ²  = 0.05, indicating that intentional learning more strongly affected memory for landmarks passed straight than for those indicating turns. We refrain from further interpretation of this interaction, however, because it is of little interest for active navigation effects.

Similar tasks have been used as measures of survey knowledge (van Asselen et al., 2006), which correlated with established survey knowledge tasks such as landmark pointing (von Stülpnagel & Steffens, 2012). However, the tour-integration task resembles a route knowledge task in many regards, as it can be solved by recalling the consecutive order of differently shaped rooms. Thus, in the present experimental setting, an advantage in tour integration performance represents an advantage in the ability to reconstruct a specific route (i.e., route knowledge) rather than in development of a cognitive map (i.e., survey knowledge).