Recent cuts in educational research funding have underscored the importance of reliable evidence on what works in education. Since 2002, the What Works Clearinghouse (WWC) has provided a resource on what works in education by reviewing causal research on educational interventions and assessing its quality. However, the WWC has prioritized internal validity (whether changes in outcomes are due to the intervention) over external validity (whether findings can be generalized to other contexts, populations, or outcomes). While this approach strengthens causal claims, it may also limit the applicability of findings beyond the studied settings. To better understand how generalizable WWC evidence is across different contexts, a 2023 study by Betsy Wolf examined the student populations and settings represented in WWC-reviewed research.
To explore this, Wolf created an evidence gap map (EGM), a tool that visualizes the distribution of existing research and identifies areas needing further study. The EGM highlights areas with abundant research and those needing more high-quality studies. An EGM typically organizes study data into a grid, where point size reflects the number of studies and color indicates study quality, offering insights to guide future research and policy decisions. The EGM results revealed disparities in the representation of school types, grade levels, and student demographics in the WWC evidence base. Specifically, WWC-reviewed studies underrepresent private schools and early childhood grades, while public schools in coastal and urban areas are overrepresented. Although student samples generally align with U.S. demographics, some groups are over- or underrepresented.
The WWC evidence base is strongest in mathematics and literacy, with less coverage of other subjects such as science, social-emotional learning, and educator outcomes. Wolf also noted that missing data on student and setting characteristics limits the generalizability of findings. The WWC’s method of assigning evidence tiers favors narrow domains with researcher-created measures over broader ones with standardized measures, raising doubts about replication. That being said, the WWC has been one of the most widely-recognized resources for educational evidence. This article highlights critical gaps in the research landscape and emphasizes the need for broader, more representative studies.
Source: Wolf, B. (2025). What works for whom: Exploring the students, settings, and outcomes in What Works Clearinghouse study data. Journal of Research on Educational Effectiveness, 0(0), 1–26. https://doi.org/10.1080/19345747.2024.2427762