A student who has been told “no” by eleven potential customers has learned something else entirely. Only one of those two changes what they do on Monday.
This is the oldest distinction in entrepreneurship education and still the most useful audit tool available: teaching entrepreneurship is not the same as teaching about entrepreneurship. Teaching about entrepreneurship explains concepts. Teaching entrepreneurship develops the ability to do it. Most programmes claim the second and deliver the first, because the first is far easier to schedule, grade and defend.
What the research says about entrepreneurship education and intention
The honest position is that the evidence for classroom entrepreneurship education is thinner than the sector likes to admit.
The most-cited meta-analysis in the field, by Bae, Qian, Miao and Fiet, pooled 73 studies covering 37,285 individuals. It found a significant but small correlation between entrepreneurship education and entrepreneurial intentions. Then it did the thing that most programme evaluations avoid: it controlled for entrepreneurial intentions that existed before the education. After that control, the relationship between entrepreneurship education and post-education intention was no longer statistically significant.
Read that carefully, because it is easy to over-claim in both directions. It does not prove entrepreneurship education fails. It shows that a large part of the measured effect can be explained by who signs up. Entrepreneurially inclined people enrol in entrepreneurship courses. The course then takes credit for the inclination.
A methodological critique published in the Journal of Small Business Management pushed the same point harder: much of the field runs on self-selected samples, no control groups, and intention as a stand-in for behaviour - because intention is cheap to measure with a questionnaire and behaviour is expensive to track for five years.
So the field measures the thing that is easy to measure, then reports it as impact. Any educator who has ever filled in a funding report knows exactly why this happens.
There is a second reason this matters now. Vocabulary-based teaching just lost its last practical excuse, because AI produces the vocabulary artefacts for free. If a course’s output is a document that a language model can generate in ten minutes, the course was never teaching entrepreneurship. It was teaching the paperwork around it.
How to redesign a course so the behaviour is the assignment
The alternative is not new and it is not mysterious. It is the experiential loop, and it looks like this:
Learn → apply → test → get feedback → fail or learn → adapt → apply again.
Compare that with the loop most institutions actually run:
Learn → exam → certificate.
The second loop can produce excellent recall. It cannot produce judgement under uncertainty, because judgement is built by making decisions with consequences and then living with the result.
Concretely, a course moves from “about” to “how” through substitutions like these:
Replace the persona worksheet with ten real customer interviews and a written record of what was actually said, not what the student hoped to hear.
Replace the sales module with one real sales conversation, prepared, conducted and debriefed.
Replace the product specification with a working prototype that someone outside the course has used.
Replace the market-sizing exercise with a pricing test - a real offer, at a real price, to a real buyer.
Replace the reflective essay with a log of assumptions killed, with the evidence that killed them.
The strongest institutional format for this is the venture creation programme, where students build an actual venture as the vehicle of the course rather than as an optional extension. Research on learning pathways within these programmes shows how much of the developmental work happens through the venture itself rather than around it. Systematic reviews of experiential approaches in university entrepreneurship education point the same way.
Now the caveat, because pretending it away is what makes this kind of argument useless. Experiential formats are expensive. They need mentors, not just lecturers. They resist standardised grading. They scale badly. They produce uneven outcomes that look messy in an annual report. Every one of those constraints is real, and each one is a reason institutions default to lectures.
But the trade-off should at least be made consciously. A programme that chooses lectures for capacity reasons is making an operational decision. A programme that chooses lectures and then reports entrepreneurial capability as the outcome is making a different kind of claim.
There is one question that separates the two, and it is worth asking of every course in a portfolio:
At the end of this course, can the student do something they could not do before?
Not define. Not describe. Do.
Knowing what an MVP is has never built one.
