Skip Navigation

Comparing Impact Findings from Design-Based and Model-Based Methods: An Empirical Investigation

This report compares empirical results from different approaches to analyzing data from randomized controlled trials (RCTs). It focuses on how impact estimates compare between recently-developed design-based methods and traditional model-based methods. Design-based methods use the potential outcomes framework and known features of study designs to connect statistical methods to the building blocks of causal inference. They differ from model-based methods that have commonly been used in education research, including hierarchical linear model (HLM) methods and robust cluster standard error (RCSE) methods for clustered designs. This study re-analyzes nine past RCTs in the education area using both design- and model-based methods. The study finds that model-based and design-based methods yield very similar impact estimates and levels of statistical significance, especially when the underlying analytic assumptions (e.g., weights used to aggregate clusters and blocks) are aligned.

PDF File View, download, and print the report as a PDF file (3 MB)