The reliability of data collected from sketch maps

Abstract Asking individuals to draw sketch maps has been a frequently used method to externalize individuals' environmental knowledge. However, nearly all studies which have utilized sketch map data have only asked subjects to draw a map on a single occasion. The implicit assumption of these studies is that the same subject will consistently produce the same map—i.e. that sketch maps are a reliable source of data, but this has never been tested. If sketch maps are not a reliable method of data collection, this would cast doubt on those studies which have used this method. The present experiment examined the test-retest reliability of sketch maps drawn by subjects who were asked to sketch the same route on two trials separated by a week. Three measures of the maps' informational content were made and all these correlated highly across the two trials, and the high correlations were maintained whether the experimental instructions were varied between trials and irrespective of how familiar the subjects were with the route which they were asked to draw. In addition, two independent judges were able to identify maps drawn by the same subject. The results demonstrate that sketch maps are a reliable method of data collection.