Exploring Pie Charts and Part-To-Whole Alternatives: Eye-tracking Approach

In the field of business information visualization (BIV), there are numerous options for displaying data. One popular yet debated choice is the use of pie charts, as their effectiveness and efficiency have been called into question. Despite their widespread use, there is limited empirical research on the effectiveness and efficiency of pie charts compared to other display types. Additionally, existing research primarily focuses on measuring user impact through accuracy and time, neglecting the impact on users’ physiological attention and cognitive resources. This study aims to fill this gap by comparing the use of pie charts to seven alternative data representation types in a part-to-whole task The study employed a randomized, within-subject experiment with 21 participants, utilizing eye-tracking technology to evaluate user performance (accuracy and time) and cognitive effort (eye fixation). The results showed that pie charts were more accurate than bar charts but required more cognitive effort than stacked bars and treemaps. Donut charts required the most cognitive effort among all tested data representations. The study highlighted that time to complete the task may not be the best indicator of user experience and accuracy while elevating the importance of accounting for users’ cognitive resources.