Expanding Frontiers in Research Designs, Methods, and Measurement in Support of Evidence-Based Practice in Early Childhood Special Education

Using Propensity Score Weighting to Reduce Selection Bias in Large-Scale Data Sets

“…There has been growing support and training for using large-scale data sets from entities such as the Institute of Education Sciences in the Department of Education, the American Educational Research Association, and the Office of Planning, Research, and Evaluation in the Office of Administration for Children and Families in the Department of Health and Human Services. Researchers in the field of early childhood studies, including early intervention, have begun to use large-scale data sets to (a) explore assessment frameworks or measurement issues in early care and education settings (e.g., Bishop, Snyder, & Algina, 2018; Gordon, Fujimoto, Kaestner, Korenman, & Abner, 2013), (b) identify groups of individuals who share similarities in measures of particular variables in specific areas by using person-oriented methodological approaches (e.g., Cook, Roggman, & D’zatko, 2012; McLaughlin, Snyder, & Algina, 2015); and (c) examine correlations between characteristics of children’s early care and education experiences and their developmental and learning outcomes (e.g., Burchinal, Peisner-Feinberg, Bryant, & Clifford, 2000; Peisner-Feinberg et al, 2001). To a lesser extent, particularly in early intervention, researchers in early childhood have begun to apply quasi-experimental methods in secondary analyses of large-scale data sets to examine plausible causal relationships between variations in children’s early learning experiences and their developmental outcomes (e.g., Ruzek, Burchinal, Farkas, & Duncan, 2014; Sullivan & Field, 2013).…”