CONVEYING INFORMATION WITH MAPS: A FUNCTION OF SYMBOLOGY

Spatial data models are often represented using common cartographic schemes which may not be the most effective for conveying information. For instance, map viewer experience and understanding may be influenced by the way continuous-value data within a map is symbolized. The impact of cartographic design characteristics has long been considered in disciplines specific to map design or cartography, but much is left to be desired for map and image design across the natural and spatial sciences as a whole. Effective symbology is pertinent for conveying continuous-value data, but there also exists a need to balance efficacy with map aesthetics in order to effectively communicate across various audiences. Similarly, Edward Tufte emphasized the importance of symbolizing data using operative color schemes: “…avoiding catastrophe becomes the first principle in bringing color to information: Above all, do no harm” (Envisioning Information by Edward Tufte, 1990). Geoscience educators, for example, often use illustrated depictions for conveying information to their students, but can neglect cartographic design principles which may limit image efficacy or worse yet, distract the image viewer.

In this paired study, we used eye tracking and Amazon Mechanical Turk to illuminate participant ability to estimate map values from a continuous-value dataset as a function of color ramp used to symbolize the data. In Part 1 of this study, we carried out an eye tracking experiment in which we quantified differences in apparent visual attention of experts and novices across images symbolized using varying common color palettes. Eye tracking data from these free gaze sessions indicated that participants interact differently when viewing continuous-value datasets symbolized using different ramps. In Part 2 of this study, we used the Amazon Mechanical Turk internet crowdsourcing tool to examine participants’ ability to estimate map values based on the color ramp used to symbolize continuous-value map data. Amazon Mechanical Turk results indicated that participant estimation of map values is strongly related to the color ramp used to symbolize the data. As a product of this study, we hope to better guide symbology selection so as to produce more effective maps for conveying information.