Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Reading in a noisy environment: An ERP study of the effects of cross-modal divided attention on prediction and repair
Poster Session E, Sunday, September 14, 11:00 am - 12:30 pm, Field House
This poster is part of the Sandbox Series.
A.S. Allevato1, K.D. Federmeier2, N.Y.Y. Wicha1; 1University of Texas at San Antonio, 2The University of Illinois Urbana-Champaign
Our brains efficiently process rapid and complex streams of input that we experience during language comprehension. We rely on sustained attention to update semantic memory as sentence context builds, allowing us to revise our expectations about upcoming linguistic information. Previous research has shown that this efficient use of sentence context during reading is reduced with divided attention, specifically when imposing a sustained secondary visual3 task or attentional load though a secondary auditory2 task. The current study aims to understand the effect of sustained auditory background noise (unintelligible babble) on the use of sentence context during reading. We measure event-related potentials reflecting access to semantic memory1: the N400, which is modulated by degree of facilitation for a specific word. We also examine the late frontal positivity (LFP), which is typically elicited during prediction violations (when a word is provided in a sentence where the context favored a different word4,5,6). Previous research using dual-task paradigms had participants respond to tones2 prior to target words or track visual distractors3 (respond to the direction of moving dots) while reading. Both paradigms revealed reduced N400s to words in sentence context when attentional resources were taxed compared to when no secondary task was imposed. The current study uses a novel cross-modal design with passive background noise while reading to investigate if prediction (N400) and revision (LFP) are modulated by sustained, cross-modal divided attention. By eliminating the secondary task we can better isolate the effects of sustained attention on reading comprehension; the cross-modal design measures attentional effects without competing for the same sensory resources. Participants read sentences from Hubbard and Federmeier (2021)4,3 that are either strongly (SC) or weakly (WC) constraining, and end with an expected (E) or unexpected (U) plausible word. The sentences are presented one word at a time either in silence or with continuous babble playing through headphones. Participants read the sentences for comprehension and reply to occasional memory probes about the sentence content to ensure engagement. No secondary task is imposed with the noise, however tones embedded in the babble can be used as individual measures of attention to the noise based on N1/P2 amplitude. This study is in the data collection phase; no statistical analyses have been conducted. Visual inspection of preliminary data (n=20) suggests a replication of Hubbard and Federmeier (2021) with graded N400 amplitude based on constraint and expectancy for reading in silence (smallest for SCE then WCE, WCU, and largest for SCU) and an LFP effect of expectancy with U targets eliciting a more positive amplitude. For trials with imposed noise, an apparent attenuated N400 effect is visible with no visible difference between WCE and WCU. Additionally in noise, we see an LFP with the strong constraint unexpected endings elicit a more positive amplitude compared to the weak constraint expected and unexpected amplitudes. These findings will inform how sustained attention affects reading comprehension in a noisy environment, like coffee shops, as well as individual differences in this ability.
Topic Areas: Reading, Multisensory or Sensorimotor Integration