Eye movements when reading novel words in text: The effect of reading purpose

Eye movements when reading novel words in text: The effect of reading purpose

First Author: Dawna Duff -- University of Pittsburgh, Communication Sciences and Disorders
Additional authors/chairs: 
Meghan Davidson
Keywords: Word Learning, Eye movements, Vocabulary, children, coherence
Abstract / Summary: 

Background: This study investigates reader goal (text comprehension, learning new words) on eye movements when reading novel words in text. Novel words create a coherence break that could trigger reader-initiated cognitive processes to resolve. Reader-initiated processes may be moderated by reader goal (van den Broek & Holder, 2017). The effect of reader goal on reader initiated processes when reading novel words is unknown.We predict reader goal will affect eye movements associated with reader-initiated but not passive cognitive processes.
Methods: We used a within subject design to compare participants reading of two passages twice-once with instructions for comprehension and once with instructions for new word learning. An EyeLink 1000 recorded eye movements with novel words as areas of interest. We fit linear mixed effects models to evaluate the effect of reading goal on eye movements associated with reader-initiated (regressions in, regressions out, regression path duration, dwell time) and passive (first fixation duration, gaze duration) cognitive processes with participant and item as random effects.
Results: Models with a fixed factor of reading goal showed a reliable effect for dwell time (b= 0.10, SE= 0.05, t=2.1) and regressions in (b=0.23,SE=0.08, t=2.96), such that both were higher for the word learning than comprehension conditions.Models with first fixation duration, gaze duration, regressions out, and regression path duration as dependent variables did not show a reliable effect of reading goal.
Conclusions: Instructions to learn new words increased reader-initiated processes in total looking time and looks back to novel words, compared to instructions to understand the text. Other eye movements associated with reader-initiated processes didn't differ with reading goal, nor did eye movements associated with passive cognitive processes.This evidence suggests that reading goal alters reader-initiated processing when reading novel words in text.