Extracting business information from graphs: An eye tracking experiment

Abstract

Information graphics are visualizations that convey information about data trends and distributions. Data visualization and the application of graphs is increasingly important in business decision making, for instance, in big data analysis. However, relatively little information exists about how people extract information from graphs and how the framing of the graphic design defines may ‘nudge’ and bias decision making. As a contribution to fill this gap, this study applies the methodology of experimental economics to the analysis of graph reading and processing to extract underlying information. Specifically, the study presents the results of an experiment whose baseline treatment includes graphical and numerical information. The authors analyze how the information extraction changes in other treatments after removing the numerical information. The experiment applies eye-tracking technology to uncover subtle cognitive processing stages that are otherwise difficult to observe in visualization evaluation studies. The conclusions of the study establish patterns in the process of graph analysis to optimize data visualization for business and policy decision making.