Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Reading Frequency Predicts Activation of Mental Imagery Networks Better Than Reading Comprehension
Poster Session B, Wednesday, September 30, 4:30 - 6:30 pm, Wangari Maathai
Michael K. Gaffney1; 1George Mason University
A substantial amount of research exists to support a correlation between mental imagery skill, often measured through behavioral assessments such as object rotation tasks, and reading skill, often measured through behavioral reading comprehension tasks. There is much less research into the relationship between how much one reads, rather than how well, and mental imagery skill. Here, I investigate this question using data from the open dataset Reading Brain Project. This project includes a reading behavior questionnaire (RBQ), whole-brain fMRI data taken from subjects who are reading technical (non-narrative) pieces, and reading comprehension scores from the Gray Silent Reading Tests (GSRT). The RBQ contains several questions that relate to reading frequency, a subset of which I selected through confirmatory factor analysis (CFI = 1.00, RMSEA = 0.00, SRMR = .070) to generate a reading frequency (RF) score. Then, after preprocessing with fMRIPrep, I used Nilearn to perform a first-level fit on the effect size of reading vs rest portions across the 5 reading tasks in the study for each of the 40 participants who had answered the necessary questions from the RBQ, had sat for the GSRT, and whose fMRI data was complete and in readable BIDS format. After identifying 13 bilateral regions from the Harvard-Oxford atlas commonly associated with creating and processing mental imagery, I performed linear regressions in R with RF scores and age-normalized GSRT quotients as predictors and the subjects’ fitted scans masked by each of the 26 regions as outcomes. The RF scores showed significant, positive associations with reading-related activation in 17 of these regions (16 surviving Benjamini-Hochberg [BH] FDR correction). These included the bilateral precuneus cortex (L: p=.012, q= .076, β=.385; R: p=.032, q=.076, β=.340) for constructing scenes, creating mental images, and placing oneself in relation to those images. They also included the bilateral superior lateral occipital cortex (L: p=.001, q=.076, β=.403; R: p= .024, q=.076, β=.357), and the left inferior lateral occipital cortex (p=.021, q=.076, β=.365), all involved in recognition of objects but not specifically of words. Complementing the right precuneus role in self-related mental imagery is the bilateral superior parietal lobule’s (L: p=.011, q=.076 β=.398, R: p=.035, q=.076, β=.334) role in spatial orientation. The subjects’ GSRT quotients, on the other hand, had significant (p<=.1), positive associations with only seven regions associated with object recognition, including upstream locations in the primary visual cortex. No significant associations with well-established imagery and spatial location regions were observed at my threshold and none survived BH FDR correction. The only region associated (by raw p-value) with GSRT strongly indicating object recognition was the left superior lateral occipital cortex (p=.047, β=.316). These findings suggest that the mere habit of reading is relevant to at least one broadly important mental skill, creating and locating oneself within mental images. Further research involving both more sophisticated measures of subjects’ reading frequency and more targeted—and reading-unrelated—tasks is needed to better understand the role of reading in the honing of a variety of critical mental skills.
Topic Areas: Reading,