From Satisfaction to Enterprise Skills: Mapping Course Evaluation onto the EU EntreComp Framework

The short version: end-of-module evaluation overwhelmingly measures reaction — did students enjoy it, was it well organised — and almost never measures whether the course developed the transferable competences graduates are hired for. The European Commission already publishes a reference framework for one crucial family of those competences: EntreComp, the Entrepreneurship Competence Framework (Bacigalupo et al., 2016), which defines entrepreneurship not as "starting a business" but as the capacity to turn ideas into value — spotting opportunities, coping with ambiguity, taking initiative, mobilising resources, learning through experience. You can map evaluation items onto EntreComp to gather indirect evidence of whether a programme builds enterprise skills. Done naïvely, with a few "I developed initiative" Likert items, it produces weak self-report noise. Done well — eliciting concrete evidence of the competence in action — it turns evaluation into something an accreditor, an employer panel, and a programme director can all use.

The satisfaction gap

Nearly all course evaluation sits at the bottom rung of Kirkpatrick's classic hierarchy: Level 1, reaction. It captures whether students were satisfied, not whether they learned, and certainly not whether they can now do anything they could not before. That would be a defensible place to stop if satisfaction were a proxy for capability. It is not: the multisection meta-analytic evidence finds that student ratings explain at most about 1% of the variance in actual learning (Uttl et al., 2017). A high overall-satisfaction score tells a programme director almost nothing about whether graduates leave more enterprising, adaptable, or opportunity-aware than they arrived.

That silence matters because the competences employers most consistently ask for — initiative, dealing with uncertainty, resourcefulness, self-efficacy — are exactly the ones a satisfaction survey cannot see. This is the same gap we have written about in moving beyond satisfaction to measuring skill development and in closing the loop on the graduate skills gap. EntreComp gives that ambition a shared vocabulary.

What EntreComp is

EntreComp was built and tested by the European Commission's Joint Research Centre and published in 2016 as EntreComp: The Entrepreneurship Competence Framework (Bacigalupo, Kampylis, Punie & Van den Brande, EUR 27939 EN) (JRC publication). Its architecture is precise and, usefully for evaluation, granular:

• 3 competence areas — Ideas and opportunities, Resources, and Into action.
• 15 competences, five in each area — including "Spotting opportunities," "Creativity," "Vision," "Valuing ideas," "Self-awareness and self-efficacy," "Mobilising resources," "Financial and economic literacy," "Coping with uncertainty, ambiguity and risk," "Working with others," "Taking the initiative," and "Learning through experience."
• An 8-level progression model and a catalogue of 442 learning outcomes, so a competence can be pitched at anything from a first-year seminar to a capstone.

Two features make it the right tool. First, it is deliberately transversal: EntreComp defines entrepreneurship as value creation of any kind — cultural, social, civic, as well as commercial — so it is no more alien to a history module (spotting an untold archival story; mobilising a community exhibition) than to a business one. Second, it is an official EU reference framework, sitting alongside the digital-competence and sustainability-competence frameworks we have mapped before in DigComp and GreenComp, and interoperable with the ESCO skills taxonomy. Evidence framed in EntreComp terms is legible to accreditors and employers, not just to an internal committee.

How to map evaluation onto EntreComp — well

The wrong way is to bolt fifteen "Rate how much this course developed your creativity (1-5)" items onto an already-fatiguing form. That inflates survey length, depresses response rates, and collects exactly the kind of low-validity self-rating the next section warns about.

The better way follows three principles:

1. Select, don't cover. Pick the two or three EntreComp competences a module is genuinely designed to develop. A field-based module might target "Coping with uncertainty, ambiguity and risk," "Taking the initiative," and "Learning through experience." Mapping to the whole framework at once is meaningless.
2. Elicit evidence, not self-ratings. Instead of "how much did you develop initiative," ask students to describe a moment when they had to take the initiative, cope with ambiguity, or mobilise a resource. A concrete narrative is far stronger evidence of a competence than a number, and it can be mapped to an EntreComp competence and progression level after the fact.
3. Frame gain honestly. Where you do use scaled items, a retrospective then-and-now framing ("compared with the start of the module...") at least anchors perceived development, though it does not escape the validity problems below.

The strongest counterarguments, taken seriously

"Self-reported competence gain is weak evidence." This is the central and correct objection. Students are unreliable judges of their own competence — the least skilled often overestimate most, and self-ratings are shaped by the reference group a student happens to compare themselves against. We have made this case at length in our piece on the validity of student self-assessment. The honest framing is therefore strict: EntreComp-mapped evaluation is an indirect, triangulating signal, never a direct measure of competence. It should sit beside direct evidence — assessed project work, portfolios, employer or placement feedback — not replace it. Claiming a survey "proves" competence development would be exactly the overreach a rigorous audience should reject.

"This instrumentalises education into a skills factory." A fair worry. Reducing a philosophy degree to a checklist of enterprise competences would be a genuine loss. The defensible position is that EntreComp mapping is additive: it captures one dimension of value a course creates, alongside — not instead of — disciplinary depth, formation, and the things that resist measurement entirely. An evaluation system can hold space for both, and should be explicit that the map is not the territory.

"Entrepreneurship is irrelevant to most of our curriculum." This dissolves once EntreComp is read as value creation rather than venture creation. Its own design intent is transversal, spanning civic and cultural value. If it still feels forced for a given module, that is a signal the module was not designed to develop those competences — which is itself a useful finding, not a failure of the framework.

"It just makes the survey longer." Only if you implement it as extra Likert items on a static form. Eliciting one narrative through conversation, rather than fifteen ratings through a grid, is the design move that avoids the fatigue and non-response penalties.

Where a conversational, AI-native platform fits

EntreComp mapping lives or dies on the quality of evidence — and narrative evidence at scale is precisely what static survey tools cannot collect and human-coded interviews cannot afford. This is Koji for Education's core strength.

• Eliciting the story, not the score. Koji's AI-moderated conversational interviews can ask a student to recount a moment of initiative or ambiguity and probe follow-up detail — producing the concrete evidence EntreComp needs, from every respondent, not just the handful a human interviewer could reach.
• Mapping narratives to competences automatically. Koji's automatic thematic analysis clusters open-text responses, so student stories can be organised against selected EntreComp competences and reported at programme level — turning hundreds of anecdotes into a defensible competence map. It goes well beyond what a sentiment score can offer.
• Structured where structure helps. Koji's six question types — including scale, ranking, and single- and multiple-choice alongside open-ended — let you combine a retrospective self-rating with the narrative that contextualises it, then weight them by response quality.
• Accreditation-ready reporting. Because EntreComp is an EU reference framework, programme-level EntreComp evidence produced through Koji speaks directly to the graduate-competence expectations that accreditation increasingly asks programmes to demonstrate — GDPR/AVG-compliant and EU-appropriate throughout.

Institutions that gather employer or alumni input on the same competences can run those studies through the main Koji platform with the identical evidence-first approach, closing the loop between what a course claims to build and what the labour market confirms.

The takeaway

Satisfaction data will tell you whether students liked a module. It will not tell you whether the module made them more enterprising, adaptable, or resourceful — the qualities Europe's own competence frameworks, and employers, actually prize. EntreComp gives you a shared, accreditation-legible way to ask that question, provided you gather evidence rather than self-flattery and treat the result as one triangulating signal among several. See how Koji for Education captures competence evidence in students' own words.

Koji surfaces indirect, narrative evidence of competence development; it complements direct assessment of learning, and does not replace it.