Risk and Resources: Early Childhood Quality and Relations to Outcomes for Children with Sociodemographic Risk

Background: The quality of care in early childhood education settings has become a primary focus for improving school readiness and development, with recent estimates demonstrating that two-thirds of all four-year-old children participate in early childhood education, up almost 25 percent from three decades ago (Magnuson, Ruhm, & Waldfogel, 2007). Currently, there is substantial heterogeneity in the quality of care across contexts, which is particularly a concern for low-income children who are more likely to benefit from, yet less likely to be exposed to, high quality care. Thus, policymakers have become increasingly interested in understanding and improving the features of care that play the most important role in children's development, and in particular, the extent to which indicators of quality serve as a facilitator or impediment for the riskiest children. One of the most widely used measures of quality in early childhood education settings is the Early Childhood Environmental Rating Scale-Revised (ECERS-R; Harms, Clifford & Cryer, 2005). Recently, the use of ECERS-R has proliferated in a number of high-stakes early childhood policies, which often include financial incentives or improvement supports that are directly tied to performance on the ECERS-R.

Methods: The current study employs data from the Early Childhood Longitudinal Study- Birth Cohort (ECLS-B) to examine the linear and nonlinear relations between ECERS-R and children's outcomes in a nationally representative dataset. Additionally, we examine the extent to which the relation is stronger for children with sociodemographic risk. We calculate children's cumulative sociodemographic risk by summing a subset of risk items (e.g., low SES, large household size, and receiving food stamps) given the recognition that multiple risk factors tend to cluster within individuals. In order to account for non-random selection into child care, we use the rich longitudinal information in the ECLS-B dataset to control for a multitude of children's individual and family characteristics from infancy to preschool.

Results: Findings from OLS regressions indicate few linear relations between ECERS-R and child outcomes. We find that higher quality on the ECERS-R benefits reading outcomes for children with sociodemographic risk. Findings are corroborated in nonlinear models, where children with sociodemographic risk significantly benefit from medium and high quality classrooms compared to low quality classrooms. Effect sizes are generally modest (range 0.03-0.07) and there is no evidence for interactions between risk and classroom quality for math, social or language skills.

Implications: Our findings suggest that the ECERS-R may not assess the components of quality that matter most for children. Although we do find benefits of classroom quality for children with sociodemographic risk in reading skills, the lack of consistency across other outcomes suggest that there may be other more predictive components of the classroom quality. Future work should continue to examine the most effective way to assess and improve quality, particularly for the neediest children.