Detecting data visualization preferences using games

Abstract

In visualizations of large multivariate data sets, discrete data can be effectively represented using glyphs. Glyphs have the advantage of allowing for rapid visual comparison, using differing visual dimensions to represent the different variables in the data. Some types of glyphs accommodate even more variables by using shape to represent the data. Yet the characteristics of these shapes may have underlying perceptual meanings. The purpose of this study was to determine whether certain shape characteristics are commonly viewed as good or bad. We conducted a study using two methods to gather data: a traditional survey, and a casual game. The results of this study strongly suggest that there are certain shape characteristics that are generally perceived as positive/negative, although they are not necessarily what might be expected.