استخراج اطلاعات تجاری از نمودارها: آزمایش ردیابی چشم  Extracting business information from graphs: An eye tracking experiment

1. Introduction Data visualization commonly helps to inform data users in a wide range of thematic areas, from health behavior to traffic accidents. Specifically, data visualization and the application of graphs are increasingly important in business decision making, policy making, and scientific research. Despite of the importance and almost universal application of data graphs and statistical charts, relatively little information exists on how people actually extract the information from graphical representations and how the framing of the graphic design defines may bias decision making and persuade users to a specific decision (Feeney, Hola, Liversedge, Findlay, & Metcalf, 2000; Lewandowsky & Spence, 1989). As a contribution to fill this gap, this study presents the results of an eye-tracking economic experiment to analyze how human beings actually scan and process statistical graphs to extract their underlying information. The application of the economic experimental methodology, including economic incentives for a successful completion of tasks, increases the validity of the results in comparison to the results from surveys or psychological experiments (Hernandez & Vila, 2014; Holt & Laury, 2002). The study helps to answer a twofold research question. From a descriptive viewpoint, the study provides experiment-based insights on how real people look at a statistical chart and which are the pieces of information that they actually process and integrate in their decision-making process. The research also studies how different visualization patterns lead to more or less convenient decisions. From a normative viewpoint, this research provides guidelines to facilitate a more effective analysis of basic statistical charts. Answering these descriptive and normative questions has not only a scientific interest—the enhancement of the understanding of the cognitive processes founding active reading of statistical charts—but also a clear managerial implication, providing with an experimental sound contribution on how to improve statistical charts to lead to better decision making in business and policy