By Emily Raich, Child Policy Certificate student ‘22
On April 13, 2021, the Center for Child and Family Policy hosted Tyler Watts, assistant professor of developmental psychology at Teachers College, Columbia University, as part of its Early Childhood Initiative Lecture Series. Watts studies educational policies designed to promote the cognitive and socio-emotional development of children from underserved communities. His talk, “Exploring Heterogeneity across Multiple Cluster Randomized Trials in Early Childhood: Evidence on intervention implementation and fadeout,” focused on his current research evaluating the Building Blocks preschool mathematics curriculum.
Developed by Doug Clements and Julie Sarama, Building Blocks takes an evidence-based learning trajectories approach to train teachers to mathematize children’s everyday experiences in the preschool classroom rather than focus on lengthy instructional time in mathematics. Building Blocks has been empirically tested by a number of scale-up studies and has been implemented in early childhood education (ECE) centers across the country.
Randomized control trials (RCTs) evaluating the efficacy of Building Blocks yield mixed results regarding the fadeout and persistence of the intervention impacts from pre-k to kindergarten. Watts’ research seeks to understand the heterogeneity across these RCTs by using a unique methodology to merge the findings from multiple studies. He explained the overarching goal of merging these results is to reveal patterns that can account for the heterogeneous outcomes across sites, thereby informing the strategies and conditions under which to most effectively scale empirically-evaluated curricula. To examine this, Watts used a methodological approach based on the work of Weiss et al. (2017) to synthesize across multiple studies and assess impact heterogeneity in ECE.
Watts is currently working with data from four scale-up RCT studies across five cities. These studies include the TRIAD study in Boston and Buffalo (2006-2007), a multi-site RCT in Nashville (2007-2008), a study in San Diego (2010-2011), and the Making Pre-K Count study in New York City (2014-2015). His study design uses a blocking method, meaning that schools with similar characteristics were grouped together in blocks within which they were randomly assigned to the treatment or control group. This method generated 47 blocks that included 4,164 children in 390 classes across 175 schools. The key component of this study design is the random assignment that occurs between the blocks and schools. School-level factors and classroom conditions—such as the time spent on math, the quality of instruction, demographics, etc.—vary across sites, and the blocking method can reveal any site characteristics that may explain the heterogeneous fadeout patterns in Building Blocks outcomes observed between pre-k and kindergarten.
The results of this study are still in their preliminary stages and models of fadeout across sites are still being investigated. These results have the potential to enhance understanding about which conditions are favorable for implementing preschool curricula. Watts’ work has important implications for efforts to successfully scale empirically-evaluated curricula and is central to ongoing discussions regarding the fadeout and persistence of early childhood intervention impacts within the field of ECE curricula evaluation.