The Science of Early Childhood Development

• In Event: Defending Childhood: Toward a Broader View of Early Education

Abstract

1. Objectives or purposes and 2. Perspective(s) or theoretical framework
In recent years, an increasing awareness of differences in children’s early learning and behavior at school entry has led policy makers at the state and federal level to turn to early childhood education (ECE) as a possible means to prevent and remediate such disparities. To date, hundreds of studies have been conducted evaluating a wide-range of center-based programs for young children. However, important questions remain about what makes a program effective. More specifically, we still know very little about the optimal timing and duration of early childhood programs (i.e., is more/ earlier better?). The impact of providing children and their families with additional support services is also not well-understood. Finally, we know very little about how program effects vary among subgroups (e.g., gender) or how early childhood education programs affect important behaviors/skills such as externalizing behavior and executive functioning.

3. Methods, techniques, or modes of inquiry and 4. Data sources, evidence, objects, or materials
In an effort to address these critical and policy-relevant questions, the National Forum on Early Childhood Policy and Programs at Harvard University’s Center on the Developing Child has constructed a comprehensive meta-analytic database of early childhood education program evaluations published between 1960 and 2007. A meta-analytic approach capitalizes on variation across program studies to identify malleable features of programs. To create this database, 10 graduate students, with guidance from four principal investigators at three universities, reviewed more then 10,000 studies of early childhood education interventions and coded approximately 300 studies that met our methodological criteria. Data from these studies were translated into more than 17,000 Hedge’s G effect sizes for inclusion in our database and analyses.

5. Results
Our preliminary analyses revealed few systematic differences by age at which children begin attending early childhood programs though longer duration programs generated more positive impacts than shorter-lived programs. We find that among family support programs, only those that provide parents with modeling of appropriate parent /child interactions and/or explicit opportunities to practice these interactions have significant impacts on children’s cognitive and pre-academic skills. Similarly, we find that only those programs that focus explicitly on improving behavior seem to be effective at preventing antisocial and aggressive behaviors. Our preliminary analyses also suggest that boys benefit significantly more than girls (one-third of a standard deviation) on important school outcomes such as grade retention and special education classification. In contrast, most of the benefits (.15 standard deviations) on adult outcomes accrue to girls.

6. Significance
This presentation will provide an overview of the database creation process and discuss the usefulness of meta-analytic studies for informing education policies and practices. Findings from results across several analyses that address the questions above will also be presented.

Authors

• Todd Grindal, Harvard University

• Jack Shonkoff, Harvard University