Geometrical distortions in geographical cognitive maps

Abstract

A map sketching methodology was applied in three studies to assess geometrical distortions in large-scale cognitive geographical maps. In the first study, 198 participants sketched Italy's national and regional borders marking the 20 regional capital locations. Furthermore, in 40 city pairs, participants had to indicate which city was further north or further west. Results showed a remarkable rotation of Italy's longitudinal axis of 26.74° toward the vertical axis. Furthermore, Italy's latitude span was significantly underestimated while longitude span was largely overestimated, leading to a more compact shape. Italy's longitude-axis rotation led to misattributions in the evaluation of longitude and latitude differences between city pairs. In the second study 185 participants were requested to sketch Sardinia, an Italy's island facing the south part of the west coastline. Results showed that in the cognitive map the island was shifted significantly toward north, facing the central part of the west coastline. As occurred for the Italy's shape in the first experiment, the island was underestimated in height and overestimated in width, resulting in a more compact shape. In the third study, 212 participants sketched the border outlines of 10 European Union countries, Great Britain and Switzerland marking the position of the national capitals. The results showed a general distance underestimation of continental capitals and a distance overestimation for London and Dublin. The capitals tended to be aligned along the vertical and horizontal axis. Overall, the three studies showed remarkable effects of rotation, alignment, shape compactness, regression toward the mean in large-scale geographical cognitive maps.