We found that for the land-use maps, the number of these cross-saccades was significantly smaller than for the two other map types. When comparing two maps of the same type, saccades between the two maps provide information on the subject’s decision-making process. At the same time, land-use maps required more fixations than the other two types of maps, while the duration of these fixations was not different from the other map types. Land-use maps required the longest response times indicating that they were most difficult to compare. Response times, fixation durations and fixation counts differed significantly for the three map types.

Subjects were asked whether two maps presented on the screen were similar (tests A and B) or different (test C). We used three types of maps that differed in their spatial granularity: (A) randomly generated, 64-by-64 pixel, black-and-white images, (B) grayscale choropleth maps representing socio-economic variables for counties in lower Michigan, and (C) land-use maps of the surroundings of selected Canadian cities in different years resulting from classified satellite imagery. Two groups of subjects with different levels of expertise with map reading were shown pairs of maps and asked to judge their similarity or difference. We propose a cognitive approach using eye movement recording to understand the process of comparing two static maps displayed simultaneously on a screen.

The process of map comparison has been studied occasionally by cartographers since the 1970s, but recent improvements in neuropsychological testing equipment and in geographical information system (GIS) technology had us review this topic in a new light. Comparing maps of different geographical phenomena, or maps of the same geographical phenomenon at different points in time, is a frequent task in many disciplines.

Mu cartographer download