Collaborative Research: Linking saccadic behavior to cognitive-linguistic processing in goal-directed reading via coupled analysis of eye movements and EEG

Reading to learn is a cornerstone of academic success and a springboard to a host of positive life outcomes. Individuals who can read efficiently enjoy better employment opportunities and therefore increased upward economic mobility, easier access to public information and therefore improved health outcomes, and greater social connectedness and civic engagement and therefore better mental wellbeing.

Students and workplace employees often resort to “skimming” a text to move through it quickly rather than carefully reading, and this increased speed often comes with a cost to comprehension. With a better understanding of the cognitive and brain mechanisms underlying reading, and how these mechanisms change under different reading goals (e.g., casual reading, skimming, reading to learn), teachers and education policy-makers will be in a better position to suggest scientifically validated ways to help every reader meet their full potential.

Our current understanding of the cognitive processes involved in reading is based on tools such as eye tracking (ET) technology or event-related brain potentials (ERPs) via electroencephalography (EEG). A major gap in understanding reading grows out of a theoretical and empirical gap between the use of these two tools. ET provides a measure of when and where the eyes move, but does not provide the continuous measure of cognitive activity that occurs before, during, and after word perception that ERPs do.

ERPs, in contrast, largely ignore reading behavior by traditionally using unnatural reading paradigms that limit and alter normal reading behavior. Emerging technologies allow for the combination of both of these tools via simultaneous measurement of eye tracking and ERP data, allowing for an expanded understanding of the relationship between brain activity and behavior in natural and goal-directed reading.

Specifically, the project investigates whether eye movements during reading are initiated prior to a word being fully recognized and whether readers use different sources of information to maintain reading speed under different kinds of reading tasks. Continuous measures of ongoing language processing directly link them to eye movements using behavior-contingent analysis techniques – that is directly examining how brain activity changes as a function of different reading behaviors, such as whether a reader decides to skip a particular word or look back to an earlier part of the text (Objective 1), relying on well-established ERP components, such as the N400 and LPC (Late Positive Component), which have been associated with post-eye-fixation language processes such as expectation and semantic integration, respectively.

In addition, the project includes exploratory analyses of saccade-related potentials in order to investigate the brain activity leading up to the movements of the eye. The project also tests the hypothesis that, relative to casual reading, different task goals will alter the relationship between brain activity and behavior so that the alignment between these processes becomes either weaker (e.g., skimming) or stronger (e.g., proofreading, reading to learn).

Therefore, this project applies the novel analysis approaches from objective 1 to gain a direct view into how cognitive processes change in response to reading goals independently from, or in coordination with, changes in eye movement behavior (Objective 2). The broader impacts of the project involve cross-training researchers with expertise in both ET and ERP approaches to reading, and an understanding of how reading works under different task goals, like when STEM students skim texts vs. reading for memorization.

This project is jointly funded by the Perception, Action and Cognition Program (PAC) and EDU Core Research (ECR).

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.