Discourse Processing in Neurologically Healthy Adults: A Cross-Modal Eye tracking Study

Abstract

One of the methods to study discourse comprehension is inference revision. Inferencing can be defined as a controlled, information-generation process that requires deductive reasoning abilities (Kintsch, 1998). Studies reported in the literature have focused on investigating inference generation during sentence processing using priming paradigms. Eye tracking tasks have been used for priming research methods to study language processing (Odekar et al., 2009). Typical eye tracking priming tasks have included participants attending to the visual stimuli by looking at the computer screen naturally, without performing any of the overt tasks (Odekar et al., 2009). Eye tracking measures offer greater temporal resolution for capturing priming effects as eye movements are recorded in real time while participants are engaged in the task. The purpose of this study is: (1) to evaluate differences in discourse comprehension abilities in younger and older groups using eye tracking; and (2) to study the relationship between discourse processing and WM abilities in both groups. Ten participants from each group participated in this study. All participants were administered language and cognitive assessment tasks to screen for general language and cognitive impairments and estimate their naming, auditory comprehension, lexical processing, cognitive, episodic memory, and working memory abilities. A cross-modal eye-tracking while listening paradigm was used to study inference revision processing. For aim 1, multiple paired sample t-tests were conducted to study the fixation duration differences for target word and control word in lexical decision task (LDT 1) and revised inference and initial inference words in LDT 2 for both groups. Further, mixed ANOVA models were run to examine the group differences (young vs. older) for LDT1 and LDT 2 tasks using four dependent eye movement measures. Question response accuracies between groups were measured using independent samples t-tests. For aim 2, Pearson correlations were obtained among working memory scores and question response accuracy scores for each group. The results showed greater fixation durations on the target word than the control word in LDT 1; and revised inference than the initial inference in LDT 2. No significant correlations were seen for working memory scores and questions developed from the revised inference sentence pair.