A growing body of literature highlights the importance of children?s early science learning opportunities for their longer-run STEM persistence, achievement, and outcomes (e.g., National Research Council, 2005; Oppermann et al., 2018). The National Research Council?s landmark report Taking Science to School (2007) highlighted the importance of providing children with science learning and inquiry opportunities in the youngest grades. In recognition of the importance of early science, preschool and pre-kindergarten standards have recently included science as a central component of children?s school readiness. The Head Start Early Learning Outcomes Framework includes scientific reasoning as a key domain of preschool children?s cognitive development, identifying specific goals and indicators for preschoolers? scientific inquiry, reasoning, and problem-solving (U.S. Department of Health and Human Services, 2015).
Despite its importance, however, children often spend very little time on science in early childhood classrooms (e.g., Brenneman et al., 2009; Early et al., 2010), and parents express less confidence that they can help their young children learn science than other subjects (Silander et al., 2018). Until relatively recently, a paucity of experimental research investigating the impacts of early childhood science interventions has hampered the field?s progress towards understanding how early science interventions? efficacy may vary by design and contextual conditions (Greenfield, 2015), leading to difficulties accumulating knowledge about educational interventions that link to children?s learning and growth in early science.
This is the first study to our knowledge to conduct a thorough meta-analysis of the rigorous causal research on the impacts of early science interventions on direct assessments of children?s science learning, as well as the quality of early science classroom instruction.
We address the following research questions:
1. What is the causal effect of early childhood science interventions on student learning?;
2. What is the causal effect of early science interventions on teacher knowledge and classroom instruction?; and
3. What malleable design elements and contextual factors of early childhood science programs predict stronger classroom instruction and student learning effects?
Via a preliminary scoping review, we have identified 28 relevant impact evaluations, most of which are randomized trials and most funded by NSF and/or IES. This number is in the same range as Slavin et al.?s (2014) influential meta-analysis of elementary science programs (k = 23); none of our identified studies were included in Slavin et al.?s meta-analysis. We expect our full search procedure to yield additional studies. As funders have increasingly prioritized early childhood science and causal impact studies, the data indicates there are now enough studies either published or in the dissemination pipeline to accumulate insights aggregated across these studies in a structured way.
Specifically, we begin by conducting a comprehensive search of both the published and grey literatures in order to capture all relevant studies that used causal designs to evaluate the impact of early science interventions. Next, the researchers will code all studies based on important features of the intervention and setting, and extract quantitative information necessary to compute effect size estimates of the impacts of the intervention on student and teacher outcomes (including student learning outcomes, social-emotional and behavioral outcomes, and measures of teachers? knowledge and classroom instruction). We will next use statistical methods to produce an overall causal estimate of the average impact of early childhood science interventions on student learning, on teacher knowledge, and on classroom instruction. Further, we will investigate how intervention effects vary by specific design elements and contextual features. The overal
University of Connecticut
07/15/2022 to 06/30/2025