Determining Spoken Language Prognosis in Children With Developmental Disabilities

Abstract

The purpose of this study is to predict which of 58 children (mean age=22 months) with developmental disabilities in the prelinguistic period of development would begin speaking 12 months after initial assessment. None of the children had severe or profound motor impairments. During the initial assessment period, children participated in a structured and unstructured communication sample with a project staff member. Also, at the time they entered the study, a mother-child interaction session was conducted to measure maternal responses to child communication acts, and mothers filled out a vocabulary checklist. Twelve months later, the structured and unstructured language samples were repeated. We labeled children with fewer than 5 different nonimitative spoken words in either communication samples as "prefunctional speakers" and those with 5 or more words in either sample as "functional speakers." The results indicate that functional speakers scored significantly higher than prefunctional speakers on 5 variables measured during the initial assessment period: (a) number of canonical vocal communication acts, b) number of intentional communication acts, c) rate of proto-declaratives, (d) ratio of words produced to those understood on the CDI (i.e., CDI discrepancy ratio), and (e) the number of maternal responses to child communication acts. After statistically controlling for the other significant predictors, only three of these variables continued to predict who would become functional speakers and who would not a year later: (a) number of canonical vocal communication acts, (b) rate of proto-declaratives, and c) CDI discrepancy ratio. These three variables discriminated functional speakers from prefunctional speakers with 83% accuracy.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 48 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.