Finland Doesn't Accredit Your Programmes or Rank You. That Changes What Course Evaluation Is For.

Finland does not accredit your degree programmes, and it does not rank your university. Its national quality agency, FINEEC, audits whole institutions on one question — are you good at improving yourselves? — and hands successful ones a quality label, not a score. Strip away accreditation verdicts and league tables, and something becomes visible that accountability-driven systems keep hidden: a satisfaction mean, the workhorse of most course evaluation, is almost useless for the only job Finland leaves on the table, which is genuine improvement.

If nobody is grading your programme and nobody is ranking you, "our average went from 4.1 to 4.2" is not a result. It is noise dressed as progress.

The model: audit, not accreditation

The Finnish Education Evaluation Centre (FINEEC, or Karvi in Finnish) runs higher-education quality assurance through institution-wide audits, thematic evaluations, and — as a voluntary, fee-based service — EUR-ACE accreditation for engineering programmes. The defining choices are these:

• No general programme accreditation. Finland does not require each degree to pass an external accreditation verdict. The audit examines the whole institution's quality system, resting on "the autonomy of higher education institutions and responsibility for developing their own activities."
• Enhancement-led by design. FINEEC describes this explicitly in a paper titled "Enhancement-led evaluation as Finland's choice". The purpose of evaluation is development, not judgement.
• A label, not a ranking. An institution that passes receives a quality label valid for six years. There is no comparative score, no ordering, no table.

This is a high-trust model. It assumes universities want to improve and that the regulator's job is to verify the capacity to improve rather than to police outputs.

Why this exposes the satisfaction mean

Most course-evaluation practice is quietly built for accountability. A single averaged number exists to be compared — this instructor against that one, this year against last, this department against the benchmark. That is what a mean is good for: ranking and monitoring. It is a summative device.

Finland removes the thing the mean is good for. No ranking, no accreditation pass/fail — so the comparative number has no external audience. What remains is the formative purpose: how do we make this course better next time? And for that purpose, a mean is close to worthless, because it does not contain the information you would act on. "Assessment scored 3.4" does not tell a teacher what to change. It is a symptom with no diagnosis.

An enhancement-led system therefore demands evaluation evidence that a satisfaction survey structurally cannot provide:

• Specifics, not scores. Not "the module was unclear" but which concept, at what point, and why — the actionable detail a Likert item discards.
• The distribution and the silence. A comfortable average can hide a bimodal split or the students who did not respond at all; non-response is often "exit," not neutrality, and enhancement needs to hear precisely the people a low-response survey loses.
• Evidence of the loop. A model built for development is judged on whether feedback produced change — the "you said, we did" trail — which a filed spreadsheet of means does not create.

Finland's design, in other words, quietly indicts the instrument most universities use. It keeps evaluation and throws away the only job the mean can do.

But isn't this just Scotland's enhancement-led review?

A fair challenge — Scotland's ELIR model is also enhancement-led, and both sit under the European Standards and Guidelines. If the argument were merely "enhancement-led review changes what evaluation is for," it would not be specific to Finland.

But the Finnish case is stronger, not just similar, because it removes more of the accountability scaffolding. Scottish institutions still operate inside a UK system with the TEF, published outcome metrics, and vigorous league tables; the enhancement ethos coexists with intense external comparison. Finland comes closer to the pure case: no programme accreditation, no national ranking culture, an audit that yields a binary label. The confound — "maybe the mean is still doing quiet accountability work elsewhere" — is weaker in Finland than almost anywhere in Europe. That makes it the cleanest natural experiment we have in what course evaluation is for once you stop using it to compare people.

The honest limitation runs the other way. A high-trust, low-comparison model relies on institutions genuinely wanting to improve and having the capacity to act. Strip out external pressure and a complacent department can coast; enhancement-led evaluation assumes good faith that accountability systems try to compel. Finland's answer is that the audit checks the improvement system itself — but that only works if the evidence flowing through that system is rich enough to drive real change. A trust-based model with thin evidence is just unexamined complacency.

Where Koji fits: evidence built for improvement, not comparison

If your evaluation exists to improve teaching rather than to rank it, you need the opposite of a mean. You need the specifics, the distribution, the silent students, and a documented loop. That is what Koji for Education is designed to produce.

Koji replaces the static Likert battery with AI-moderated conversational interviews that ask why and follow up for the actionable detail — converting a 3.4 into "the second assignment's brief contradicted the rubric, and it surfaced too late to recover." Automatic thematic analysis turns programme-scale open text into named, ranked issues a course team can act on; quality scoring protects the signal from careless responses; and because collection can run formatively and mid-cycle, a problem can be fixed this term rather than noted for the next six-year audit. Action tracking and programme-level reporting build the "you said, we did" evidence trail that an enhancement-led audit is actually looking for, and the standardised, bias-aware AI moderator keeps that evidence consistent across departments without human-moderator drift. Koji surfaces and helps you act on what students mean; it does not claim to eliminate bias, and no responsible tool does.

Finland made the honest choice: evaluation is for getting better, not for keeping score. The instrument most universities still use was built for keeping score. That is the mismatch worth fixing. (Teams doing general user or staff research can use the same interview engine on the main Koji platform.)

Running an enhancement-led evaluation and tired of means that tell you nothing to change? See how Koji for Education produces improvement-ready evidence.

Frequently asked questions

How does FINEEC evaluate higher education in Finland?

FINEEC (Karvi) audits whole institutions on the effectiveness of their quality systems, runs thematic evaluations, and offers voluntary EUR-ACE accreditation for engineering. It does not impose general programme accreditation and does not rank institutions; successful ones receive a quality label valid for six years.

What does "enhancement-led evaluation" mean?

It means the purpose of evaluation is development rather than judgement or accountability. FINEEC describes this as "Finland's choice": the model starts from institutional autonomy and responsibility for improving one's own activities, not from external verdicts.

Why does an enhancement-led model make the satisfaction mean useless?

A mean is a summative device built for comparison — ranking instructors, years, or departments. Finland removes ranking and accreditation, leaving only the formative question "how do we improve this course?" — for which a mean contains no actionable information.

Isn't this the same as Scotland's ELIR?

Both are enhancement-led and ESG-aligned, but Finland removes more accountability scaffolding: no programme accreditation and no national ranking culture, whereas Scotland's model coexists with the TEF and league tables. Finland is the cleaner case of evaluation-for-improvement.

How does Koji fit an enhancement-led approach?

Koji produces improvement-ready evidence: conversational interviews that yield the specifics behind a rating, thematic analysis of open text, mid-cycle collection so problems are fixed within a term, and action tracking that documents the loop an enhancement-led audit examines.