Symbol Considerations for Bivariate Thematic Mapping

Abstract

Bivariate thematic maps are powerful tools for understanding geographic phenomena, making visible spatial associations between them. But bivariate thematic maps are more visually complex than a univariate map, a source of frustration for both map creators and map readers. Despite a variety of visual solutions for bivariate mapping, there exists few 'best practices' for selecting or implementing an appropriate bivariate map type for a given scenario. This results in a need for empirical research examining the perceptual and functional differences among bivariate mapping solutions. This research reports on a controlled experiment informed by the theory of selective attention, a concept describing the human capacity to tune out unwanted stimuli, and attend specifically to the information desired. The goal of this research was to examine if and how the perceptual characteristics of bivariate map types impact the ability of map readers to extract information from different bivariate map types. 55 participants completed a controlled experiment in which they had to answer close ended questions using bivariate maps. Accuracy and response time were recorded for each question. The experiment also opened with biographical questions to determine participant expertise and finished with a Likertbased survey to determine participant preference of the different map types. The results of this experiment suggest that 1) despite longstanding hesitations regarding the utility of bivariate maps, participants were largely successful in extracting information from most if not all of the tested map types, and 2) the eight map types varied in terms of how intuitively participants were able to use the maps to answer the survey questions. While selective attention theory could explain some of these differences, the performance of the map types differed appreciably from similar studies that examined these map symbols in a more abstract, perception-focused setting. While the perceptual models of selective attention can still be useful in guiding map design, more work needs to be done in understanding the cognitive aspects and limitations to bivariate map reading.