What Matters When Providing Conversational Recast Treatment? A Multilevel Modeling Analysis

Area of Science:

• Speech and Language Pathology
• Developmental Psychology
• Linguistics

Background:

• Conversational recasting is a common intervention for children with developmental language disorder (DLD).
• Variations in treatment delivery and target selection may influence effectiveness.
• Enhanced conversational recasting (ECR) incorporates linguistic variability and attentional cues.

Purpose of the Study:

• To directly compare multiple variations of conversational recasting.
• To determine how modifications in delivery and target impact treatment outcomes for children with DLD.
• To assess the relative difficulty of learning different morphological targets.

Main Methods:

• Bayesian hierarchical linear mixed modeling was used.
• Outcome data from 141 children with DLD across 10 recast treatment versions were compared to ECR.
• Data from 132 children treated for one of five morphological targets were analyzed.

Main Results:

• Seven of 10 treatment conditions showed decreased generalization probe performance compared to ECR.
• These conditions resulted in 47%-84% worse performance than ECR alone.
• One morphological target was easier to remediate; the others showed no difference in acquisition difficulty.

Conclusions:

• Treatment delivery variations can hinder or facilitate learning in children with DLD.
• Clinicians must carefully consider treatment structure for optimal outcomes.
• Target selection may offer more flexibility in conversational recasting interventions.