Sequential Spatial Choices in the Large-Scale Environment

Abstract

Sequential spatial choices of shopping destinations in the largescale environment, entailed by multistop trip planning, have been modeled as a sequence of choices of closest locations. The validity of such modeling attempts was questioned by the results of a reported study. In a laboratory test, subjects who solved simulated trip-planning problems with one, two, or three choice sets with pairs of familiar locations were found frequently to choose in a way that total distance was minimized even though this meant that the closest location was not chosen each time. The results were similar in a field test in which subjects were taken to a downtown business district, asked to choose locations, and to walk to the chosen locations. Subjects were furthermore found to make few revisions when implementing their choices. The ability to minimize total distance probably relied on the discovery of spatial configurations of locations that appeared to presuppose a simultaneous representation of the locations, as suggested by differences in results in both the laboratory and the field tests depending on whether subjects had access to a street map or not, in the laboratory test depending on whether the locations were downtown or all over the city.