Comments of the Future of Evidence in Education Network on Proposed Revisions to EDGAR

Introduction

The Future of Evidence in Education Network is a group of researchers and evaluators working to strengthen how education evidence is generated, interpreted, and used. Our members bring expertise in research methods, evaluation, evidence standards, education policy and research and development (R&D), and the use of evidence in state and local decision-making. We submit these comments on the Department of Education’s proposed revisions to the Education Department General Administrative Regulations (EDGAR; 91 Fed. Reg. 54666 (Aug. 24, 2026)).

Our comments draw on two recent Network reports developed through a convening of researchers, evaluators, and evidence leaders: Necessary but Not Sufficient: Six Design Principles for a Stronger Education R&D System and Evaluating Causal Evidence in Education: A Practical Guide. Necessary but Not Sufficient identifies six public functions needed to produce evidence that is rigorous, transparent, cumulative, and useful across settings and over time. Evaluating Causal Evidence in Education presents ten principles for judging what causal evidence can support and whether it is relevant and complete enough to inform a particular decision.

We are writing to express concern about three proposed changes that we believe would weaken the rigor, consistency, and independence of federally supported evidence in education: (1) new provisions concerning grant termination and continuation; (2) the proposed definition of “strong evidence”; and (3) the proposal to remove the What Works Clearinghouse (WWC) Handbooks as the required common benchmark for strong and moderate evidence determinations while retaining them as only one optional review pathway.

1. Continuation and termination provisions should protect the independence and completion of federally supported research.

Current EDGAR provides structured rules for continuing multiyear awards and incorporates government-wide procedures for terminating awards. Section 75.253 gives continuation awards priority over new grants when recipients satisfy requirements related to progress, reporting, eligibility, financial and administrative management, and the Federal Government’s best interest. It also requires notice of the grounds for denying continuation and an opportunity to request reconsideration. Section 75.901 currently cross-references the government-wide rules governing termination, notice, objections, hearings, and appeals. Those rules allow termination for noncompliance, by mutual agreement or at the recipient’s request, or under clearly stated award terms. The rules also require written notice and establish procedures for objections, hearings, and appeals. (Current EDGAR, 34 C.F.R. §§ 75.253 and 75.901; 2 C.F.R. §§ 200.340–200.342 (2026))

The proposal would allow an administration to reduce or terminate otherwise compliant grants that no longer align with its policy priorities. Proposed § 75.253 would end the priority for continuation awards and permit funding to be delayed or released partially or in installments. Proposed § 75.901 would allow termination “for convenience of the Secretary or pass-through entity.” Together, these changes expand the government’s discretion to withdraw support after an award begins (Proposed rule, §§ 75.253 and 75.901).

Ending studies early because agency priorities change can waste federal investments and deprive state, district, and school leaders of evidence needed for consequential decisions. Multiyear research depends on continuity: researchers must complete planned data collection and analyses and report all findings, including those that conflict with current policy preferences. Making continued support contingent on shifting political priorities would weaken scientific integrity, reduce the public return on federal funding, and make federal research support less reliable and more partisan. The risk of abrupt termination could also deter researchers and institutions from undertaking ambitious, long-term studies.

Necessary but Not Sufficient recognizes the legitimate role of officials in setting broad research priorities. But it also recommends safeguards that preserve researchers’ authority to carry out planned analyses and report findings that potentially conflict with policy priorities or sponsor interests. (Necessary but Not Sufficient, pp. 10-11.) Tying research funding (and results) to the political winds undermines the scientific integrity of research.

We recommend retaining the current EDGAR provisions governing continuation awards and not adding a general termination-for-convenience provision to § 75.901. The final rule should also state that a change in administration-wide policy priorities, standing alone, is not sufficient grounds to reduce or deny continuation of, or terminate otherwise compliant research that is making substantial progress.

2. A statistically significant positive result from one study should not be sufficient to establish “strong evidence.”

Proposed § 77.1 would allow “strong evidence” to rest on one well-designed and well-implemented experimental study showing a statistically significant positive effect. Unlike current EDGAR, the proposed definition would not expressly require that the study population and setting overlap with those proposed to receive the intervention, that the evidence be based on multiple sites with a sufficient sample size, or that there be no overriding negative findings. The proposed definition also does not require reviewers to consider the broader body of research when multiple studies exist or assess whether the estimated effect is large enough to matter in practice. These are additional safeguards recommended in our reports, not requirements in current EDGAR.

The Network strongly supports rigorous impact studies as an essential part of the education evidence system. We do believe that a single well-conducted study can provide credible evidence that an intervention caused an effect in the population and setting studied. But a single study does not necessarily constitute a sufficiently strong and broadly applicable body of evidence to warrant a “strong evidence” designation.

Evaluating Causal Evidence in Education explains why populations and settings matter. Even the most rigorous study can show only what occurred for the people and in the places studied. Whether similar results should be expected elsewhere depends on the students served, the comparison condition, staffing and implementation supports, and other features of the local context. The report therefore recommends assessing the similarity between the studied and intended populations and settings and considering the full body of evidence when multiple studies exist. Evidence across multiple contexts can help distinguish results that are relatively stable from those that depend on particular conditions (Evaluating Causal Evidence, pp. 21-23).

Necessary but Not Sufficient argues that education leaders need cumulative knowledge about what works, for whom, and under what conditions. Understanding what works, for whom, and under what conditions requires bodies of evidence that capture meaningful variation across populations and settings and are documented in ways that support comparison and synthesis. Federal evidence definitions should advance this cumulative process, not suggest that a single favorable study defines the evidence base for an educational intervention.

Finally, the proposed definition would allow a statistically significant positive finding to support a strong-evidence designation without requiring evidence that the effect is educationally important or worth the costs and demands of implementation. Evaluating Causal Evidence explains that statistical and practical significance answer different questions for decision-makers (pp. 18-19).

We recommend that the Department retain the safeguards the proposal would remove from the current evidence-tier definitions: relevance to the intended population and setting; multi-site and sample-size requirements where applicable, and consistency in the evidence. The Department should also maintain rigorous study quality standards and require strong evidence determinations to consider the broader body of research when multiple studies exist, as well as effect magnitude, precision, and practical importance relative to implementation costs and demands. Together, these protections would prevent a context-specific, anomalous, or inconsequential positive finding from being treated as strong and broadly applicable evidence.

3. Flexibility in evidence review should not come at the expense of common, transparent, and consistently applied standards and procedures.

The proposed changes would weaken the common, transparent foundation for federal evidence designations. Current EDGAR anchors strong and moderate evidence determinations in publicly available WWC standards. The proposed framework would allow various organizations and peer review panels to apply the required evidence framework without using a common detailed protocol. Although the proposal requires the framework to address core features of study design, measurement, implementation, and analysis, it does not require every review pathway to publish how the criteria were applied in a particular case or disclose the scope and limitations of an individual determination. Although this approach offers greater flexibility, it creates a risk that the same evidence could receive different designations depending on the Department grant competition and the organization or peer reviewers assessing it.

Our reports emphasize that evidence must be reported and reviewed transparently. Necessary but Not Sufficient specifically recommends that review organizations publish their criteria, protocols, and scope; explain how the criteria were applied; and preserve sufficient information for users to understand what evidence was considered and what a strong or moderate evidence designation does and does not support (Necessary but Not Sufficient, pp. 12-13). Although the proposal makes its general framework public, it does not require all reviewers to follow comparably detailed public protocols or explain how they reached individual determinations. By expanding flexibility without these safeguards, the proposal risks making federal evidence designations less transparent, consistent, and useful to decision-makers.

By anchoring strong and moderate evidence determinations in the WWC, the current EDGAR standards are designed to ensure that studies are assessed against common evidence standards. The WWC Procedures and Standards Handbook provides a free, public resource that can be used by many different agencies and organizations to determine the quality of evidence. The Handbook has been updated over time to incorporate new knowledge and best practices in the field of education research. Application of these standards and procedures helps to ensure interventions designated as having strong or moderate evidence meet the same evidence standards. Our concern is not with allowing alternative review pathways. It is that, without comparably detailed and transparent procedures, the common criteria may be interpreted and applied differently across pathways, making the rigor of individual determinations less consistent and less verifiable. The same evidence could therefore receive different designations depending on the pathway used, creating confusion for users and opportunities to rely strategically on the pathway most likely to yield a favorable rating.

Conclusion

Greater flexibility and efficiency should not come at the expense of credible, useful, and independent federal evidence. We urge the Department to preserve continuity for compliant research and evaluation awards making substantial progress; shifting policy priorities alone should not justify reducing or terminating them. We also urge the Department to ensure that “strong evidence” signals not merely one statistically significant positive result, but study rigor, contextual relevance, scale, consistency across the available research, and practical importance. Finally, evidence reviews should follow detailed public protocols, apply standards consistently, and explain how each determination was reached and what the designation does – and does not – support. These protections are essential to a cumulative and transparent evidence base that state, district, and school leaders can understand, trust, and use.