Graduation Year

2025

Document Type

Dissertation

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Communication Sciences and Disorders

Major Professor

Ruth Huntley Bahr, Ph.D.

Committee Member

Hana Kim, Ph.D.

Committee Member

JiSook Park, Ph.D.

Committee Member

Elizabeth R. Schotter, Ph.D.

Keywords

Predictive Social Mind framework, visual world eye-tracking paradigm, prediction processing, social cognition conflict resolution, emotion word processing, facial emotion recognition

Abstract

Individuals must recognize linguistic and non-linguistic social cues to resolve contextual ambiguities in their conversations. When the linguistic and social cues match, sentence processing is facilitated. However, conflicts between linguistic and social stimuli (i.e., incongruency) can lead to prediction errors which slow sentence processing. Limited research has explored the social cognition conflict resolution skills in individuals with Developmental Language Disorder (DLD) despite their documented emotion processing deficits. This study considers the role of social prediction processing in adults with DLD as they engage in a social cognition conflict resolution task.

Young adults with DLD (n=29) and typically developing (TD) peers (n=42) participated in an eye tracking study that used the Visual-World Paradigm (VWP) to assess the impact of linguistic and non-linguistic social stimuli on real-time auditory sentence processing. Group membership was determined utilizing an established language battery for identification of individuals with DLD (Fidler et al, 2011). Using the VWP, participants first viewed a facial expression prime that was positive or negative in valence. The participants then heard an auditory sentence with an emotion word that either matched the emotional valence of the facial prime or was emotionally incongruent with the facial prime. Two picture scenes that differed in emotional valence were presented at the same time as the sentence. The participant viewed the facial prime, listened to the sentence, and selected the picture that best matched the meaning of the sentence. Independent variables included congruency (valence of facial expression prime and embedded emotion word) and language ability (DLD or TD).

Statistical analyses considered eye movement behavior, as well as reaction time (RT) and accuracy of participant responses. A generalized additive mixed effect model (GAMM) analyzed the proportion of fixations to the target picture. Data was analyzed before and after target emotion word onset. Results revealed differences in the time course of sentence processing between groups. The TD group seemed to use an adaptive strategy, prioritizing accuracy/caution over speed to resolve incongruencies. The DLD group appeared to use a suppression strategy whereby they prioritized the current linguistic aspects of the sentence. A linear mixed effects (LME) model was run to measure task accuracy and RT. Both groups had slower RTs for the incongruent condition; however, the DLD group had higher accuracy rates in the incongruent condition, which may indicate more careful and deliberate processing. Overall, these findings suggest individuals with DLD do generate social predictions based on initial social information but are slower to update their mental representation when conflicting social information is presented.

Download

Share

COinS