You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster D43, Thursday, November 9, 6:15 – 7:30 pm, Harborview and Loch Raven Ballrooms

Individual Competence in Reading Comprehension and Fluid Intelligence Modulates Right DLPFC Activity when Reading Scientific Texts

Chun-Ting Hsu1, Benjamin Schloss1, Ping Li1;1Pennsylvania State University

Reading is one of the most prominent methods to learn abstract scientific concepts in education. Reading comprehension involves a sequence of cognitive processes, including word recognition, semantic and syntactic processing, world knowledge integration and situation model updating, leading to coherence building. Although linguistically defined, these processes also heavily depend on other cognitive abilities like working memory, executive function, and attention. The competence of reading comprehension would affect the ability to acquire scientific concepts, which is also reflected in the individual measurement of fluid intelligence. Using a design enabled by naturalistic self-paced reading, this study investigated neural correlates of reading comprehension of scientific texts at the level of sentential processing, and the neural correlates reflecting individual differences in the reading comprehension competence and fluid intelligence while reading. Fifty participants read five scientific texts (math, mars exploration, electric circuit, ship building and safety, and GPS) sentence by sentence in a self-paced method inside the scanner while eye-tracking data and multiband EPI of TR = 400 ms were acquired. Participants’ reading comprehension ability (measured by the Gray Silent Reading Test; Wiederholt & Blalock, 2000) and non-verbal fluid intelligence (measured by the Raven’s Progressive Matrices; Raven et al. , 1988) were collected in their second visit. In the GLM analysis of the fMRI data, using the timing of first fixations on content words as onsets, we modeled the parametric effect of the word position index in a sentence (e.g., coding 3 indicated the third word in a sentence) of content words. We found regions involved in visual feature processing to be more active at the beginning of the sentence, which included bilateral visual cortex and left fusiform gyrus, as well as left dorsolateral prefrontal cortex (dlPFC, MNI: -48 -1 42). By contrast, the regions that were more active toward the end of the sentence included bilateral regions associated with syntactic processing (insula), modality specific semantic processing (pre- and postcentral gyri), semantic retrieval (dorsomedial prefrontal cortex, dmPFC, and supplementary motor area), coherence building (temporoparietal junction, TPJ, and dorsomedial prefrontal cortex), attention and working memory (dlPFC, MNI: -24 14 58 & 48 29 34, and TPJ). More specifically, one peak in the right dlPFC (MNI: 27 14 38) became more activated towards the end of the sentence in participants who had higher scores in reading comprehension, and another peak in the right dlPFC (MNI: 18 23 42) became more activated towards the end of the sentence in participants with higher non-verbal fluid intelligence. Our results suggest a possible early focus on surface form analysis and later focus on semantic analysis and integration during reading. DLPFC has been associated with executive function, attention, and working memory, and our finding showed right dlPFC activity towards the end of the sentence to be correlated with reading comprehension ability and fluid intelligence. They further suggested the role of dlPFC in the learning of abstract scientific concepts. Causal relationships between executive functions and reading comprehension at the behavioral and neural level should be further investigated.

Topic Area: Meaning: Discourse and Pragmatics

Back to Poster Schedule