Age-related changes in ERP components of semantic and syntactic processing in a verb final language

Introduction Several researchers suggested that older adults performed less accurately and more slowly than younger adults on sentence comprehension tasks (DeDe et al., 2004; Kemtes & Kemper, 1997; Waters & Caplan, 2001). However, it has been controversial whether there might be age-related changes during online processing. Previous research employed a word-by-word reading or listening paradigm to measure older adults’ online sentence processing abilities. In contrast, there are few studies to qualitatively examine age-related differences in how semantic and syntactic information are processed using event-related potential (ERP). The current study investigated age-related differences in offline and online sentence processing using ERP in a verb-final language. Methods Sixteen younger adults (mean age=25.8, SD=3.1) and fifteen older adults (mean age=73.0, SD=6.0) participated in the study. Ninety target stimuli consisted of three conditions: (1) plausible, (2) syntactic violation on the second noun phrase with an irrelevant case marker (Word2), (3) semantic violation on the third word containing an irrelevant verb (Word3). Each phrase (Eojeol) was presented for 700ms using a Rapid Serial Visual Presentation paradigm with a 200ms inter-stimulus interval during the 32-ch EEG recording. Results 1. Behavioral data Accuracy and response time (RT) to the plausibility judgment served as dependent measures. Two-way mixed ANOVA revealed that the main effect for the group was significant in the accuracy, (F(1,29) = 4.53, p < .05) and the RT (F(1,29) = 58.92, p < .01). No other effects were significant. 2. ERP data Two-way ANOVAs (group x condition) were performed for each midline site (Fz, Cz, and Pz). For Word2 at 300-500ms, significant main effects for condition emerged across the site (all ps. However, these effects were diminished in the elderly group only at Fz (F(2, 58)=3.801, p