Date of Completion
Honors College Thesis
Department of Communication Sciences & Disorders
Honors College, College of Arts and Science Honors
Narrative comprehension, prediction, autism, semantic processing, electroencephalography, event-related potentials
Autistic individuals have shown differences from non-autistic individuals when understanding stories, regardless of whether those stories are told through words or pictures. One aspect of narrative comprehension that is thought to facilitate comprehension is the prediction of upcoming words or events in a story. Prediction is typically assessed using cloze probabilities: the predictability of a word in a linguistic narrative, or a panel in a visual narrative. In studies measuring event-related potentials (ERPs) with non-autistic adults, more predictable (“high cloze”) words/panels elicit reduced N400 amplitudes compared to less predictable (“low cloze”) words/panels. Autistic individuals may have differences in predictive processes compared to non- autistic individuals, which could contribute to differences in narrative comprehension across modalities. Here, we report two studies using the cloze probability paradigm to investigate predictive processing mechanisms in the context of linguistic and visual narrative comprehension among autistic and non-autistic adults. Non-autistic adults showed modulations of the N400 ERP component by cloze in both linguistic and visual domains, which replicates prior literature. Autistic adults showed earlier N400 modulations by cloze in both modalities. This suggests that autistic adults employ a more bottom-up processing style during narrative comprehension which supports models that propose differences with prediction in autistic individuals.
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License.
Kubinski, Sarah, "Effects of predictability on visual and linguistic narrative comprehension in autistic and non-autistic adults" (2023). UVM Honors College Senior Theses. 563.
Available for download on Thursday, May 08, 2025