Research

Risk Factors and Services for Vocabulary Delays in Early Childhood: Population-based Estimates

Principal Investigators: Paul Morgan and Marianne Hillemeier, Penn State; George Farkas, UC-Irvine; and Carol Hammer, Temple University

Duration: Two years

Funding: Goal 1, Early Intervention and Early Learning in Special Education, National Center for Special Education Research

Purpose

This project investigates the role of potentially malleable factors (including parenting practices, child care quality, and EI/ECSE services) in the onset of vocabulary delays during at risk children’s infant, toddler, and preschool years, and consequences for school readiness. We operationalize school readiness using measures of children’s reading and mathematics achievement, as well as their learning-related, externalizing, and internalizing behaviors. We address five key questions. First, which factors most strongly predict children’s vocabulary knowledge at 24 months of age? Second, which children are most likely to receive early intervention/early childhood special education (EI/ECSE) when they are 24-48 and 48-60 months of age? Third, which children are most likely to display vocabulary delays at 48 months of age? Fourth, which factors strongly predict children’s general cognitive and behavioral functioning at 24 months, as well as their pre-academic and behavioral functioning at 48 months of age? Fifth, which children are most likely to display lower academic and/or behavioral readiness at 60 months of age? The project’s activities will help identify the extent to which increasing at risk children’s oral vocabulary knowledge at 24 and/or 48 months constitutes a possible mechanism for bolstering their academic and behavioral readiness at 60 months of age.

We also investigate the previously untested hypothesis that naturally delivered EI/ECSE services (e.g., speech/language therapy) helps increase delayed or disabled children’s cognitive, preacademic and academic, and behavioral functioning. Thus, our Goal 1 project’s activities should help identify many potentially malleable factors contributing to at-risk children’s school readiness.

Setting

We will analyze the Early Childhood Longitudinal Study-Birth Cohort (ECLS-B). These data represent children born in the U.S. in 2001 (sampled from birth certificate files), who were then assessed, and other data collected, at ages 9, 24, 48, and 60 months of age.

Sampled Population

The population of U.S.-born children of preschool age, including oversampling of low birth weight children and racial/ethnic minorities. 

Primary Research Method

Secondary data analyses of standardized direct measures and questionnaire data collected in the ECLS-B.

Measures and Key Outcomes

Measures include individually-administered and parental reports of children’s vocabulary knowledge, cognitive functioning, reading and mathematics achievement, and direct observation and teacher ratings of learning-related, externalizing, and internalizing problem behaviors. Measures also include direct observation of parent-child interactions, direct observations of childcare quality, birth certificate data on socio-demographic, gestational, and birth characteristics, and parent interviews. Outcomes include children’s vocabulary knowledge, cognitive functioning, reading and mathematics achievement, and learning-related, externalizing, and internalizing problem behaviors.

Data Analysis Strategies

These include cross-lagged models with extensive statistical control for likely confounds and weighted adjustments to account for sample clustering, multiple imputation to account for missing data, and propensity score matching methods to better identify the predicted effects of natural variation in receipt of specific EI/ECSE services on the school readiness of children with identified delays or disabilities. These analyses should provide the best available estimates of many previously hypothesized causal relations, including the relation between at-risk children’s early vocabulary knowledge and their later school readiness.


 

Secondary data analyses of standardized direct measures and questionnaire data collected in the ECLS-B.