Your Course Evaluation Was Built for 19-Year-Olds. Most European Learners Aren't.

BLUF: The typical end-of-semester course evaluation was designed around an implicit persona: an 18-to-22-year-old, full-time, on-campus student progressing through a three-year degree. That persona is increasingly the exception. As European higher education absorbs the EU's lifelong-learning agenda — micro-credentials, upskilling, returning professionals — a growing share of "students" are working adults with prior experience, defined goals, and severe time constraints. Malcolm Knowles's theory of andragogy predicts that these learners judge a course on fundamentally different criteria than school-leavers do. An instrument that never asks about those criteria will systematically mismeasure the fastest-growing segment of the sector. This piece explains where the standard survey fails adult learners, addresses the objection that "good teaching is good teaching," and sets out what andragogically valid evaluation should capture.

The scale of the shift

Adult learning is not a niche. In 2024, 28.5% of EU adults aged 25 to 64 had participated in education or training in the previous twelve months, and the EU has committed to reaching at least 60% of adults in training every year by 2030 under the European Pillar of Social Rights (Eurostat, Adult learning statistics, 2024). The 2022 Council Recommendation on a European approach to micro-credentials made short, targeted, outcome-focused learning a formal policy instrument for "lifelong learning and employability" (Council Recommendation, 16 June 2022). The direction of travel is unambiguous: universities will evaluate more and more provision aimed at adults who are not on a traditional degree track.

Yet the evaluation instrument has not moved. Most course surveys still ask a returning 45-year-old software engineer the same satisfaction and workload questions designed for a first-year undergraduate — and interpret the answers against undergraduate benchmarks.

Why andragogy breaks the standard survey

Knowles's andragogy — his framework for adult learning — rests on assumptions that differ sharply from the pedagogical model built into most course evaluations. Adults, he argued, are self-directed, bring a large reservoir of prior experience, are oriented to problems and immediate application rather than to a distant syllabus, and are motivated internally by relevance to their goals. Each assumption exposes a blind spot in the standard instrument:

• Relevance to real goals, not course coverage. A traditional survey asks whether the course was well-organised and the content clearly explained. An adult learner cares first whether it solved the problem they enrolled to solve — whether they can now do the thing at work on Monday. Standard instruments rarely ask.
• Use of prior experience. Adults judge a course partly on whether it respected and built on what they already knew, or wasted their time re-teaching it. There is no item for "did this course make use of my experience?" on a typical form.
• Immediacy and transfer. For adults, the relevant outcome horizon is short and applied. This connects directly to the transfer-of-training question — will the skill survive contact with the job — which matters far more to a mid-career learner than to a second-year student with years before graduation.
• Time and life constraints as first-class context. For a full-time student, "workload" is a demand to be measured. For a working parent, workload interacts with a job and caring responsibilities in ways that make a raw hours number nearly meaningless without context.

When the instrument asks the wrong questions, its silence is not neutral — it tells adult learners the institution is not interested in what they came for, which itself depresses response and candour.

This is a measurement-validity problem, not just a UX complaint

It would be easy to treat this as a matter of tone. It is deeper. If adult learners evaluate against a different construct — problem-solving relevance and applicability rather than delivery and coverage — then applying a delivery-focused instrument to them is a case of construct underrepresentation: the instrument omits the dimensions that matter most to the population. And benchmarking adult-learner scores against traditional-cohort norms violates measurement invariance, because the same item does not mean the same thing to the two groups. A micro-credential that scores "average" on a survey built for degree modules may be excellent as a micro-credential — a mismatch we explored in why short courses break the end-of-semester survey. Averaging across such groups produces numbers that are precise and wrong.

The counterargument: "good teaching is good teaching"

The strongest objection is that clarity, organisation, fair assessment, and expertise are universal virtues — good teaching is good teaching, so why do adults need a special instrument? Partly true, and worth conceding: the core teaching qualities do transfer, and we should not pretend adult learners live in a separate pedagogical universe.

But two things break the "one instrument fits all" conclusion. First, even shared virtues carry different weights: an adult with a specific job problem weights applicability far above breadth of coverage, and an instrument that gives every item equal implicit weight will misrank courses for them. Second, adults face additional criteria that simply do not arise for traditional students — respect for prior experience, schedule compatibility, immediate transfer — and an instrument silent on those is not "neutral," it is incomplete for that population. The claim is not that adults need a different theory of good teaching; it is that they need an instrument that measures the parts of good-teaching-for-them that the standard form leaves out.

What andragogically valid evaluation should capture

Evaluating adult and lifelong provision well means adding constructs, not just softening tone:

• Goal attainment: did the learner achieve the specific outcome they enrolled for? This is closer to a learning-outcomes-met question than to a satisfaction rating.
• Applicability and transfer: can they use it now, in their actual context? Probe for concrete examples, not a Likert agree/disagree.
• Respect for experience: did the course build on or ignore what they already brought?
• Fit with life: was pacing, scheduling, and format compatible with working life — and where did it strain?
• Formative, not just terminal: adults on short courses cannot wait for an end-of-semester survey; mid-cycle formative feedback is often the only feedback that can help them at all.

Where Koji fits

Adult learners have the least patience for a generic ten-item Likert form and the most specific things to say. Koji for Education is well-suited to this because its AI-moderated conversational interviews adapt to the respondent: instead of a fixed undergraduate questionnaire, the moderator can ask a returning professional whether the course solved their problem, where they applied it, and what their prior experience made redundant — following up in the learner's own words. The six structured question types let programmes capture goal attainment and applicability as first-class data, while automatic thematic analysis makes sense of highly heterogeneous adult feedback at scale — surfacing, for example, that career-changers valued the applied project while upskillers wanted deeper theory. Because collection is fast and can run mid-cycle, it fits the compressed timelines of micro-credentials and short courses. Koji does not claim to erase the differences between learner populations; it lets you measure them instead of flattening them into one inappropriate average.

The same need — understanding a heterogeneous population on its own terms rather than through one fixed script — is why teams doing user and customer research use the main Koji platform and its shared AI interview engine.

The takeaway

If your institution is serious about the lifelong-learning agenda, the evaluation instrument has to grow up with the learner base. A survey built for 19-year-olds will keep telling you your micro-credentials and professional courses are merely average, because it never asks the questions on which adults would rate them excellent. Add the constructs that matter to adults — goal attainment, applicability, respect for experience, fit with life — and stop benchmarking adult provision against undergraduate norms. The learners changed; the instrument should too.

Evaluating micro-credentials, executive education, or returning learners? See how Koji for Education adapts to every learner.