Gender Differences in Eye Movements During Online Reading

This paper provides evidence of individual difference on digital reading, by examining male/female and good/poor readers’ eye tracking records (including pupil, blinks, fixations, saccades, and regressions). From 6,520 undergraduate students’ university entrance exam scores and the follow-up reading assessments, we strategically selected 74 participants (half males and half females, top 3% good readers and bottom 3% poor readers) from different disciplines. Results indicated that the major differences between male and female readers were saccade duration, regression rate, and blink rate. Good and poor readers differ significantly in most of the eye tracking indicators, and the major effects of reading ability were more obvious than the gender effects. Among all the eight indicators examined in this study, blink and regression rates are most sensitive to gender attribute, while fixation rate and saccade amplitude presented least sensitiveness. These findings would be helpful for user modeling with eye-tracking data in intelligent tutoring systems.