Spatial Memory - Virtual Reality

Virtual Reality

During one recent study researchers designed three different virtual towns, each of which had its own "unique road layout and a unique set of five stores." However, the overall footprint of the different maps was exactly the same size, "80 sq. units." In this experiment, participants had to partake in two different sets of trials.

First, participants were assigned two of the three virtual landscapes and tasked with the role of a taxi driver. The participant’s avatar picked up a total of 25 passengers and dropped them off at random locations throughout the map. Researchers measured the participants’ "delivery path length" in both sets of the first trial. They found that there was a "clear decrease in path length with increased number of deliveries in a town." The participants’ improved route length correlated with the increasing amount of experience they got driving in the virtual towns. However, as one would expect, the knowledge that participants gleaned from the first town did not transfer, or aid them, in their travels around the second town. Thusly, Newman et al. (2006) inferred that the participants "formed a survey representation of each town." That is they drove around enough of the town to infer the general layout of the rest of it.

The second set of trials involved the same task for the participants but with some changes to the context of the environment. Researchers added two more maps and made them smaller. It should be noted, though, that the second map set in this trial was standardized with the same layout and landmarks, meaning that all participants would experience this map on their second run. Moreover, three of the five maps "replaced some set of the landmarks with novel landmarks while the layout of target locations remain(ed) unchanged." The other two maps were either "identical" to the second traversed landscape of the second trial or both the landmarks and their relative locations were changed. Again, Newman et al. measured "excess path length," specifically focusing on the excess distance traveled on the first delivery of the second map in the second trial, where they believed that spatial learning would transfer mostly strongly from the first map. They found that the group that encountered altered building representations in the same locations had the highest level of layout knowledge transfer from the first town.

The results of the first trial showed that people are capable of learning the spatial layout of an interactive environment. The second trial showed that despite a change in landmark presence, participants were still able to "find novel shortest routes." Because participants did not travel across previously memorized routes and still performed well according to the "excess path length" standard, Newman et al. concluded that these results evidenced "some higher-order survey representation of the environment.