Previous research demonstrates that children who spend more weekly hours in early education programs demonstrate larger cognitive gains than children who attend fewer weekly hours (Li, 2013).  Given mounting evidence of the importance of high quality early education for low-income children’s positive outcomes (e.g., Burchinal, Vandergrift, Pianta, & Mashburn, 2010), we hypothesize that this dosage effect may vary by the quality of that early education setting.  Thus, the present study explores the extent to which classroom quality moderates the impacts of hours spent in in early education on children’s language and math skills. 

This study uses data from the Head Start Impact Study, a randomized controlled trial of the effectiveness of Head Start (n=4,440).  Classroom quality was measured using the Early Childhood Environment Rating Scale (ECERS) and the Arnett Caregiver Interaction Scale (CIS).   Child outcomes include early receptive language (Peabody Picture Vocabulary Test) and math skills (Woodcock Johnson III Applied Problems) directly assessed in the spring of the preschool year.

Preliminary analyses were conducted using OLS regression to predict child outcomes from weekly hours in early education programs, the quality of that program, and their interaction, controlling for children’s pre-test scores.  Results of these analyses find a significant interaction between hours and quality:  For children attending early education programs for a small number of hours per week, quality was not related to math achievement, but for children who spent many hours per week in early education programs, attending a high quality setting was significantly associated with higher spring math skills controlling for fall skills, while attending a low quality setting was associated with lower math skills.  The interaction effect predicting children’s receptive language was not statistically significant.

Future analyses to be presented as part of this panel will explore a variety of quasi-experimental methods to improve causal inference, including an instrumental variable (IV) approach, to estimate unbiased impacts of weekly hours in early education classrooms on child outcomes and the moderating role of classroom quality.  Previous work using these same data (Li, 2013) suggests that the number of weekly hours offered by a center may be a strong instrument to assess the causal effects of hours in care on outcomes for children. Additional instruments such as random assignment to Head Start will also be explored to address the potential endogeneity of the quality of early childhood programming.  Although previous analyses indicate that both instruments significantly predict our “treatments,” we will explore the extent to which each instrument meets critical IV assumptions.

Early childhood programs are often forced to make trade-offs between offering high quality educational programs to support children’s development and full-day child care for working families. The current paper will explore the relationship between length of program day and quality.  Our very preliminary results show that both should be considered and that they should be considered in tandem (not independently).  Complete results and their policy implications will be presented as part of this panel.