Maternal and child vocabulary: Concurrent and predictive relationships from ages two to three

The purpose of this study was to investigate concurrent and predictive relationships between early maternal and child vocabulary in a group of 70 low-income, ethnic-minority families who had applied for Early Head Start services. Further, we were interested in how variation in the amount of vocabulary that mothers used related to their children's linguistic and cognitive development. Dyads were videotaped during 10-minute play sessions when children were both 24 and 36 months. Videotapes were transcribed and analyzed for mother-child vocabulary with the assistance of CHILDES. Children's linguistic development was assessed with the PPVT and the Index of Productive Syntax (IPSyn) at 36 months. The Bayley Scales of Infant development provided a measure of children's cognitive development at both 24 and 36 months.; Both mothers and children varied tremendously in the amount and diversity of vocabulary that they used at each assessment. Overall, children's performances on tests of linguistic and cognitive development were considerably below age-expected norms. However, there was also substantial variability, and some children tested within normal limits. There was a positive correlation between maternal language and children's productive vocabularies at both ages. Multiple regression analyses demonstrated that maternal language at 24 months uniquely predicted children's performance on tests of linguistic and cognitive development at 36 months, but did not predict children's productive vocabularies. Cluster analyses revealed three distinct patterns of maternal-child vocabulary use. Children from the two groups in which mothers used a high amount of vocabulary at 24 months preformed better on tests of linguistic and cognitive development at 36 months relative to a group in which mothers used little vocabulary—regardless of how much the children themselves spoke.; In conclusion, early maternal vocabulary was a significant predictor of children's early linguistic and cognitive development in this low-income sample. Some children were exposed to scant maternal language, and lagged far behind developmental norms. However, others were exposed to rich maternal language, and their development was well within normal limits. A deficit model of low-income families in which children hear and acquire little language is inadequate to capture the great variability that exists in this population.