Date of Award
Spring 2015
Project Type
Thesis
Program or Major
Psychology
Degree Name
Master of Arts
First Advisor
Edward J O'Brien
Second Advisor
Jill McGaughy
Third Advisor
Robert Ross
Abstract
In prior research, there is a consistent finding that readers do not detect semantic anomalies when they are embedded in text, especially if what they are currently reading shares a high degree of featural overlap with information previously encountered in the text. The present experiments further examined the processing of semantic anomalies embedded in narrative passages. In four experiments, the interaction of contextual information relative to a semantic anomaly and the reader’s standard of coherence were systematically manipulated to assess the impact of the anomalous information. The results demonstrated that the participants were disrupted by anomalous information, but whether this was immediate or delayed depended on the amount of supportive context. Experiments I demonstrated that when a high amount of contextual information was active in memory it delayed the disruption caused by anomalous information. Experiment II demonstrated that when the contextual information was not active in memory, comprehension was disrupted immediately. Experiment III demonstrated that raising the participant’s standard of coherence resulted in an earlier disruption in reading caused by anomalous information. Experiment IV demonstrated that lowering the participant’s standard of coherence resulted in a delayed disruption or no disruption at all. The results of all four experiments are consistent with the RI-Val view of comprehension.
Recommended Citation
Williams, Christopher Ryan, "Processing Semantic Anomalies: Validation Against General World Knowledge" (2015). Master's Theses and Capstones. 910.
https://scholars.unh.edu/thesis/910