Sentence Prediction Deficits in Developmental Language Disorder Are a Product of Vocabulary Knowledge and Processing...

Area of Science:

• Linguistics
• Developmental Psychology
• Speech and Language Pathology

Background:

• Children with developmental language disorder (DLD) exhibit challenges in predicting linguistic elements following verbs.
• Understanding the interplay between lexicosemantic knowledge and online processing is crucial for DLD research.

Purpose of the Study:

• To investigate the roles of lexicosemantic knowledge deficits and online processing deficits in sentence prediction difficulties in children with DLD.
• To compare performance on offline lexicosemantic and online processing tasks in varying sentence speeds and complexities.

Main Methods:

• Twenty-six 4- to 5-year-old children with DLD and 26 typically developing (TD) peers participated.
• Lexicosemantic knowledge was assessed via offline tasks, and online predictive processing was measured using eye tracking.
• Experiments manipulated sentence speed and complexity to evaluate processing demands.

Main Results:

• Lexicosemantic knowledge quality significantly influenced sentence prediction accuracy in both groups.
• Children with DLD demonstrated poorer lexicosemantic knowledge than TD peers.
• While prediction was similar in faster sentences, children with DLD showed prediction decay in slower sentences, even after controlling for knowledge.

Conclusions:

• Sentence prediction deficits in DLD stem from a combination of impaired lexicosemantic knowledge and processing challenges, including working memory decay and attention.
• Both knowledge-based and processing-related factors contribute to sentence prediction difficulties in developmental language disorder.