Individuals' eye movements in reading are highly consistent across time and trial.

Abstract

Eye movements are used to study a variety of cognitive phenomena, including attention, perception, memory, language, reading, decision making, and many others, as well as cognitive impairments and individual differences in cognition. These studies assume, with little evidence, that eye movements are stable across time and trials. Eye movement stability must be better characterized to understand the full theoretical and clinical implications of individual differences in eye movement behavior. The present study examined eye movement reliability in normal individuals during reading. Thirty-nine participants completed 2 sessions of a reading task separated by 1 month. Means and standard deviations of fixation duration, saccade amplitude, first fixation duration, gaze duration, total time, go-past time, skipping, refixation and regression probabilities were compared both between sessions and across trials within sessions. All correlations were highly significant, indicating that eye movement behaviors are stable within individuals across several weeks and highly stable across trials within each individual. The different components of the ex-Gaussian distribution of fixation durations were also highly stable over time. Differences in sensitivity to lexical variables (frequency, predictability, length) were also compared, and were also observed to be highly stable across time. Eye movements in reading are therefore suitable for studying cognition and its neural underpinnings, as well as cognitive development and longitudinal change. Theoretical and clinical implications of these findings are discussed. (PsycINFO Database Record