ABSTRACT
We often consult maps when we visit a city for the first time. It is frequently the case that we do not look at the map continuously, but repeatedly consult it to memorize the way to a certain point and then walk to that point. By analyzing the memorizing process for a route, we can see how to construct easy-to-comprehend road networks and where to locate road maps on the streets. Another important question is the memorizing unit for routes. Can we quantitatively analyze memory loads for route segments? To answer these questions, experiments through computer-aided walking simulation (Kamachi, 1988) have been conducted and the results analyzed.¶
A variety of studies related to walking routes have been performed in the field of spatial cognition. (For the review of studies in spatial cognition on a large spatial scale, see Siegel and White (1975).) For example, Thorndyke and Hayes-Roth (1982) analyzed knowledge acquisition from maps, and showed that maps are an effective device for revealing the relative spatial relationship. Cousins et al. (1983) analyzed children’s ability of route finding, and Passini (1980) summarized concepts for route-finding
mechanisms. Moar and Carleton (1982) also tried to analyze route-memorizing mechanisms. The existing studies largely use the actual space for their experiments. A variety of factors therefore interfere with the results, and the pure effect of those factors cannot be precisely analyzed. Even in an apparently featureless field, it was found that subjects utilized subtle differences in the actual space as landmarks (Heft, 1979). To overcome this difficulty, we experiment with completely featureless streets without landmarks on a computer. The final goal of our study is to quantify how much people can memorize each element in memorizing a route. Experiments are based on the preliminary study by Asami et al. (1988), and the results are analyzed subsequently.
