The Effects of Prediction and Speech Rate on Lexical Processing
Files
Publication or External Link
Date
Authors
Advisor
Citation
DRUM DOI
Abstract
How are we so good at quickly and accurately comprehending speech? Researchers have suggested that we not only process language as we hear it, but also predict what we are likely to hear next (e.g., if you heard "The researcher built a time machine to travel into the.." you might be expecting "future"), which helps our comprehension. What is not yet clear is the necessity of prediction accuracy (e.g., what if you predicted "past" instead of “future”?) or the cognitive processes that underlie this prediction process. The goal of this project was to investigate the effect of lexical prediction on language processing, and how this effect varies with speech rate and individual cognitive abilities. In an active prediction paradigm, participants heard two-sentence passages at fast, medium, or slow rates while predicting the final word of the passage. Instead of hearing a final word, participants were instructed to read a word aloud as quickly as possible, then indicate if this was the word they predicted. The word had about a 50% chance of matching the participant's prediction. Results show that both correct and incorrect prediction facilitated reading time, suggesting that prediction facilitates language processing regardless of prediction accuracy. Additionally, slower speech rate resulted in slower response time across conditions, indicating that language processing may slow to match speech rate. Response times were related to general processing speed, but the effects of prediction accuracy and speech rate were not related to measures of either processing speed or working memory. In all, these results support the hypothesis that active prediction decreases language processing time, which may also be affected by speech rate and general processing speed.