Authors: Nahid Ferdous, Radu Jianu
Abstract: Changes in human pupil size are known to be correlated with task difficulty. Here we explore the opportunity of using eye tracking to measure task difficulty in the specific context of data visualization. In particular, we use a controlled eye-tracking study to investigate the difference between two types of task difficulty, mental difficulty and visual difficulty, explore the time frames at which pupil size responds to changes in task difficulty, and investigate if pupil size can provide qualitative hints as to which part of a task people find difficult. We found that eye tracker reveals mental difficulty more precisely than visual difficulty. We also found a set of patterns of pupil size changes that are related with mental activity and we show that using pupil size in conjunction with gaze coordinates lets us make inferences about user cognition that would not be possible if looking at gaze coordinates alone.