When an edtech vendor claims its product is evidence-based, the claim maps onto a federal scale with four tiers — and the bottom tier, under the Every Student Succeeds Act of 2015, requires only a rationale, not a result. The tier structure is defined by the US Department of Education and applied by its Institute of Education Sciences' What Works Clearinghouse, which reviews studies against fixed design standards before counting them as evidence. A vendor citing 'research' without naming a tier is usually citing tier IV or nothing, and the difference between the tiers is the difference between a study and a theory.
This primer gives teachers and procurement staff the vocabulary the sales deck assumes you don't have. All standards cited are from the Department of Education's published ESSA evidence tiers and the Clearinghouse's procedures as of 2025.
What are the four tiers?
| Tier | ESSA name | What it requires |
|---|---|---|
| I | Strong evidence | At least one well-designed, well-implemented randomized controlled trial |
| II | Moderate evidence | At least one well-designed quasi-experimental study |
| III | Promising evidence | At least one well-designed correlational study with statistical controls |
| IV | Demonstrates a rationale | A logic model and proposed study — no result required |
The table's last row is the one to memorize. Tier IV is a plan, and it is a legitimate category — every product starts there — but a slide that says 'research-backed' while resting on tier IV is describing an intention, not a finding. The Department of Education's own guidance is explicit that the tiers rank strength of evidence, not quality of product.
Who checks the studies?
The What Works Clearinghouse does — or doesn't. The Clearinghouse reviews studies against its design standards: randomized trials need adequate equivalence and low attrition; quasi-experiments need credible comparison groups, per its published procedures handbook. A study the Clearinghouse has reviewed carries a visible rating; a study no one has reviewed carries whatever the vendor's slide says about it.
The gap is the practical issue. The Clearinghouse has reviewed a small fraction of the thousands of products in the market, so most claims live unreviewed — which doesn't make them false, but does make them the vendor's characterization of its own evidence. The house distinction applies: the company says versus the clearinghouse found.
How do you question a claim in a sales meeting?
Five questions, in order, each with one right kind of answer:
- Which ESSA tier is the claim resting on? Tier IV ends the evidence conversation; it should start a timeline one.
- Who ran the study — an independent evaluator or the vendor? Vendor-run pilots are vendor data, useful and labeled.
- What was the comparison condition? 'Gains versus the start of the year' is not 'gains versus a control classroom.'
- What population and outcome measure, published where and when?
- Has the What Works Clearinghouse or a regional education lab reviewed it?
Question three is where most decks wobble. A within-group gain proves time passed and teaching occurred; only a comparison condition separates the product from the classroom around it.
What does the record not establish?
That any given product works for your students, in your subject, at your dosage. Efficacy studies run in specific settings with specific implementations, and the Department of Education's guidance on the tiers explicitly frames evidence as a starting point for local judgment, not a substitute for it. The tier system exists to make claims checkable. The checking is still your job.
For more context, read Retrieval practice works, but not as much as trainers claim.