I Have Research on My Program. What ESSA Level Is It?

What Actually Determines a Study’s ESSA Tier?

A study’s ESSA tier is determined by its research design, sample, and statistical results — not by how large or impressive the findings are. The tiers rank how confidently a study can attribute an outcome to a program, which is a statement about methodology rather than product quality.

That distinction matters more than almost anything else in this conversation. A rigorously designed randomized trial that finds a small effect can be rated at the highest tier. A program with several studies and much larger reported gains can land at the lowest, because the designs behind those gains can’t rule out other explanations. So you aren’t asking “how good is my program?” You’re asking “what does the structure of this particular study allow anyone to claim?” The second question has a far more determinate answer.

Notice the last column of the Tier IV row. The statutory definition asks not only for a logic model but for an ongoing effort to examine the program’s effects. A rationale document filed and forgotten is weaker than the tier suggests; a rationale document paired with a study in the field is exactly what the tier describes.

How Do I Tell What ESSA Level My Study Is?

Work through three questions in order: is there a comparison beyond the students’ own starting scores, how was that comparison formed, and did the study meet the sample and significance requirements. Most research places itself within the first two.

What Disqualifies a Study From an ESSA Tier?

Six issues account for most downgraded ratings: a non-independent outcome measure, a single-site sample, an analysis that ignores clustering, undocumented baseline equivalence, results that aren’t statistically significant, and research that’s too old to describe the current product.

• The outcome measure is your own. If the only evidence of growth comes from an assessment embedded in your product, reviewers will discount it. The measure needs to be independent — a state test, a widely used benchmark assessment, or a standardized instrument the district administers for its own purposes.
• The sample is one district. Tiers I and II require a multi-site sample, generally read as at least two schools or districts and at least 350 students. A single friendly partner site, however clean the data, caps you at Tier III.
• The analysis ignores clustering. Students within a classroom aren’t independent observations. An analysis that treats them as though they are will overstate precision, and reviewers who catch it will move an otherwise qualifying study down a tier.
• Baseline equivalence was never documented. A quasi-experimental design without evidence that the groups started in comparable places isn’t really a quasi-experimental design. This is the single most common reason a study that calls itself Tier II gets rated lower.
• The findings weren’t statistically significant. Encouraging trends and directionally positive results don’t qualify for any tier. A study that finds no significant positive effect remains valuable for product development, but it isn’t evidence in the ESSA sense.
• The research is dated. Several states apply a recency window — Arizona’s guidance points toward research conducted within roughly the last fifteen years. A strong study of a product version that no longer resembles what you sell can be technically sound and practically unpersuasive.

Who Decides What ESSA Level My Research Is?

Three different parties make the call, and they don’t always agree: the vendor self-designates, an independent clearinghouse may rate the study against published standards, and state review committees apply their own rubrics and can adjust a claimed tier.

• You do, first. Vendors self-designate when responding to an RFP or submitting to a state list. This is legitimate as long as it’s honest and documented. Michigan’s Levels of Evidence Worksheet works exactly this way — vendors select a tier, submit the study, explain the design and outcomes in plain language, link to the research without a paywall, and document significance and sample details. The state’s Committee for Literacy Achievement then verifies the claim and can adjust it downward.
• Independent clearinghouses do, second. Evidence for ESSA, run out of Johns Hopkins, and the What Works Clearinghouse apply published standards and publish the result. Their ratings carry weight precisely because they aren’t yours. They are also stricter than most state rubrics on clustering, sample size, and retrospective designs.
• States do, third, and on their own terms. State requirements are not copies of the federal definitions. The same study can be Moderate evidence in one submission, Promising in another, and unrated by a clearinghouse.

Do All States Apply the Same ESSA Requirements?

No. States set their own evidence rules on top of the federal definitions, and the differences are consequential — Arizona requires a control group even at Tier III, which the federal correlational standard does not.

Why Should Someone Else Review My Study Before I Submit It?

Because knowing your tier is only half the problem. The other half is knowing which states and clearinghouses your evidence will actually satisfy, and in what order to approach them — and that requires someone tracking rules across dozens of jurisdictions that change every legislative session.

A vendor can read the federal definitions and reach a defensible conclusion about a study’s design. What’s much harder to do from inside a company is answer the questions that follow, each of which turns on knowledge of the review landscape rather than knowledge of the study:

• Which lists is this evidence actually good enough for? A study that falls just short of Evidence for ESSA’s threshold may comfortably clear several state rubrics. Knowing which ones saves you from writing off usable evidence.
• Where should it go first? A published low rating is difficult to undo. Submitting a borderline study to the strictest reviewer before strengthening it can lock in a designation you’d rather have avoided.
• When do the windows open? State adoption cycles have deadlines. Missing one can mean waiting a full procurement cycle, regardless of how good the evidence is.
• Does the sample match the state’s students? Michigan explicitly evaluates the match between your study population and Michigan students. A strong study conducted in the wrong demographic context scores lower than a modest one conducted in the right one.
• What does the portfolio need next? Most companies are better served by a cross-section of studies at varying levels than by one study pursued to perfection — so that a search on ERIC or Evidence for ESSA returns a substantive profile rather than a single entry.
• What makes the claim land? A tier designation carries more weight when it arrives with the details reviewers look for — sample size and setting, the outcome measure, significance, baseline equivalence. Documenting those turns a self-designation into something a procurement committee can verify, which is what separates an evidence claim that opens doors from one that invites questions.

This is the work LXD Research does before designing anything: read what you already have, place it accurately, and map it against the states and clearinghouses where it will do the most good. Since 2020 we’ve had more than 25 studies approved by Evidence for ESSA — 64% at the two most rigorous tiers — and supported state approval processes in Arizona, Michigan, Utah, Arkansas, and Wyoming. That experience is mostly useful to clients as pattern recognition: we’ve seen which arguments reviewers accept and which they send back.

What Do You Get From a Study Review?

Every LXD Research review produces four things you can use immediately: an educator-friendly summary written for district audiences, a named researcher’s authorship on that summary, an LXD Research ESSA badge reflecting the tier the evidence supports, and publication to our ResearchGate profile so the work is discoverable outside your own website.

The reason this matters is timing. Third-party review is slow by design — clearinghouse queues run months, and state adoption cycles open once a year. Meanwhile your sales team is answering evidence questions this quarter. The deliverables below are meant to close that gap without overstating what you have.

• An educator-friendly summary, co-branded with your product. Effect sizes translated into the terms practitioners actually use, educator voices included, benchmark context provided. It’s written to stand on its own so a rep can hand it to a curriculum director or attach it to a conference session, without requiring anyone to read a full technical report.
• Named authors on the work. Our researchers put their names and credentials on what they write. A summary signed by an outside researcher who can be looked up carries a different weight with district reviewers than an unsigned internal white paper — and 75% of educators report trusting tools more when they’re independently validated, while trusting company marketing considerably less.
• An LXD Research ESSA badge you can use right away. The badge reflects the tier your evidence supports under an independent researcher’s assessment, and it’s available as soon as the review is complete. It isn’t a substitute for an Evidence for ESSA or state designation, and we’re explicit about that distinction — it’s what you use while those reviews are pending. Our badge criteria and definitions are published, so anyone can check what a badge means.
• Publication to ResearchGate and ERIC. Completed studies are indexed where researchers, state reviewers, and procurement teams already search — more than 100 LXD Research reports now sit in ERIC and academic databases. Evidence hosted only on your own site invites the question of who produced it; evidence in an independent index doesn’t.

Taken together, that’s the difference between saying your program has evidence and being able to show someone where it lives, who wrote it, and what standard it was held to.

What If My Study Lands Lower Than I Hoped?

A lower tier is a real evidence claim, not a failure — and with roughly three quarters of widely used edtech tools carrying no ESSA-aligned evidence at all, a documented Tier III or Tier IV still puts you ahead of most of the market. There are usually three paths forward.

• Reanalyze what you already have. If the partner district has records for students who didn’t use the program, a matched comparison group may already exist inside data in hand. This is the least expensive route to a higher tier and is available more often than companies assume.
• Run a look-back study. Existing implementations and district-administered assessments can support a comparison design built retrospectively — no new data collection, no imposition on partner schools, and a workable path to Tier III or Tier II depending on how the comparison is formed.
• Treat the current study as a floor. Document the logic model so the Tier IV claim is solid, launch a prospective study on this year’s implementations, and let the evidence base build in layers.
• What won’t work: adding random assignment retroactively. Tier I always requires a new prospective study, so a Tier I ambition is a planning decision rather than an analysis decision.

That last sequence — rationale documented, promising evidence published, moderate evidence in the field — is what a mature research portfolio actually looks like. It also reads to reviewers as a company that takes evidence seriously, rather than one that produced a single study and stopped.