Course Evaluation Evidence for French Accreditation (Hceres & CTI)

In short: In the French system, student evaluation of teaching — l'évaluation des enseignements par les étudiants (EEE) — is both a legal obligation and a recurring expectation of the two main quality-assurance bodies: Hcéres (the Haut Conseil de l'évaluation de la recherche et de l'enseignement supérieur) and, for engineering programmes, the CTI (Commission des Titres d'Ingénieur). Both operate under the European Standards and Guidelines (ESG) — Hcéres and the CTI are members of ENQA and registered on EQAR. What evaluators want is not the survey itself but evidence of a closed loop: that you collect student feedback, analyse it, act on it, and can show the result. This guide maps those requirements to concrete outputs.

The legal baseline: EEE is mandatory in France

Student evaluation of teaching has a longer legal history in France than many institutions realise. It traces to the arrêté Lang of 26 May 1992, was reinforced by the arrêté Bayrou of 9 April 1997, and is set out in the arrêté of 22 January 2014 fixing the cadre national des formations for licence, licence professionnelle and master degrees. That decree requires institutions to put in place a mechanism for the evaluation of teaching by students (un dispositif d'évaluation des enseignements par les étudiants) with the explicit aim of continuous improvement of programme quality.

Two consequences follow for procurement and quality leads:

1. EEE is not optional, and "we run an end-of-term survey" is not, by itself, evidence of compliance. The decree asks for a dispositif (a system) oriented to amélioration continue (continuous improvement). Collection without action does not satisfy it.
2. Many institutions formalise this in a charte d'évaluation des enseignements setting out anonymity, how results are used, and who sees them. Evaluators will look for that governance.

What Hcéres looks for

Hcéres is an independent public authority that evaluates institutions, programmes and research units on a five-year cycle organised in vagues (waves). Its methodology is built on autoévaluation (institutional self-assessment) followed by external peer evaluation against a published référentiel. In the Wave E (2024–2025) referential for institutional evaluation, the quality of programmes and the continuous improvement of teaching sit among the structuring domains, and student participation in the quality process is expected.

For evaluation evidence, Hcéres is interested less in raw scores than in the functioning of the quality loop:

• Is there a dispositif for EEE that actually runs across programmes?
• Are results analysed and discussed in programme committees (conseils de perfectionnement)?
• Do they lead to documented changes?
• Are students involved in the process and informed of outcomes?

This is where Likert-only surveys often fall short for the audit: they produce numbers, but not the narrative of action an Hcéres panel wants to see.

What the CTI looks for (engineering schools)

For écoles d'ingénieurs, the CTI has accredited engineering programmes by law since 1934. Its expectations are published in the Références et Orientations (R&O), a multi-book framework (general criteria, a self-evaluation tool, and procedures). The CTI is explicit about a démarche qualité and continuous improvement (amélioration continue) running through the school: student feedback is one of the inputs that must feed the improvement cycle, alongside outcomes data and stakeholder input. The CTI has been a full member of ENQA since 2005 and EQAR-registered since November 2010 (renewed through 2024), so its expectations are aligned with the ESG.

For a CTI visit, evaluation evidence should show the feedback-to-action chain at both the enseignement (course) and formation (programme) level, and demonstrate that the school uses it to manage the engineering competency profile.

Mapping requirements to concrete Koji outputs

Requirement (Hcéres / CTI / ESG)	What evaluators want to see	Concrete Koji output
A functioning EEE dispositif (arrêté 2014)	Systematic collection across courses, not ad hoc	Standardised AI-moderated evaluations run every term, per course / programme
Amélioration continue / closing the loop	Evidence that feedback led to change	Themes linked to actions, with status tracked cohort to cohort
Analysis discussed in conseils de perfectionnement	Readable thematic syntheses, not raw exports	Automated thematic analysis with prevalence and representative quotes
Student involvement and feedback to students	"You said / we did" communication	Loop-closing summaries that can be shared back
Longitudinal view across the accreditation cycle	Trend over the five-year vague	Cohort-over-cohort theme and score tracking
ESG Part 1 (student-centred, ongoing monitoring)	Standardised, fair process	Consistent neutral AI moderation across all cohorts
GDPR / data protection	Lawful, anonymised handling	EU/GDPR-focused data handling and anonymity by design

The point is not that a tool produces accreditation for you — it does not. It is that standardised, analysed, action-tracked evidence is far easier to present to an Hcéres or CTI panel than a folder of PDF survey exports.

Turning feedback into audit-ready evidence: a practical sequence

1. Collect systematically. Run EEE every term, at enseignement and formation level, with a consistent instrument so cohorts are comparable.
2. Probe, don't just rate. Conversational follow-ups capture the reason behind a judgement — the material a conseil de perfectionnement can actually act on. (On why open text beats Likert-only, see what open-text comments reveal.)
3. Analyse into themes. Replace manual reading with thematic analysis so each programme has a short, evidenced synthesis.
4. Act and record. Tie each theme to an action and an owner; record what changed.
5. Close the loop with students. Communicate "you said / we did" — both an ESG expectation and a response-rate booster.
6. Aggregate for the wave. Maintain a cohort-over-cohort trend so the self-assessment dossier shows a managed cycle, not a one-off snapshot.

Where a traditional survey tool may still suffice

In fairness: if your institution already runs a mature EEE process — strong participation, active conseils de perfectionnement, and robust manual analysis capacity — a conventional survey platform (EvaSys, Explorance Blue, Qualtrics or similar) can meet the letter of the requirement, especially where the priority is a stable quantitative series for internal benchmarking. The gap AI-moderated evaluation closes is qualitative depth and the labour of turning comments into documented action — precisely the part Hcéres and CTI panels scrutinise. If that part already works for you, the marginal benefit is smaller.

Common evidence gaps French panels flag

Across Hcéres and CTI visits, the same shortcomings recur — and most are not about the survey instrument but about what happens after collection:

• Numbers without narrative. A dossier of mean scores with no account of what was discussed or decided. Panels read this as collection, not a quality loop.
• No trace of the conseil de perfectionnement. EEE results should visibly enter the programme committee's deliberations; minutes that reference specific feedback and the resulting decisions are strong evidence.
• Open-ended comments left unanalysed. Hundreds of free-text answers exported to a spreadsheet and never coded. This is exactly where automated thematic analysis turns a liability into an asset.
• Low or unexamined response rates. A 15% response set presented without any discussion of representativeness invites challenge. Acknowledge non-response, and show how conversational depth compensates for lower counts.
• No feedback to students. The "you said / we did" step is frequently missing; it is both an ESG expectation and the single most effective way to rebuild participation.
• No longitudinal view. A single snapshot rather than a cohort-over-cohort trend across the vague makes it hard to demonstrate a managed, improving cycle.

Closing these gaps is less about buying software and more about treating evaluation as a documented decision process. A platform helps by making the analysis and action-tracking automatic, so the evidence assembles itself as you go rather than in a scramble before the visit.

How this fits the wider European picture

France's framework is one national implementation of the same ESG logic that underpins evaluation evidence across Europe. If you operate across borders, see our companion guides for the Netherlands and Flanders (NVAO), the UK (TEF and QAA), Germany (Programm- and Systemakkreditierung), business-school accreditation (AACSB and EQUIS), and the overarching ESG / ENQA evidence guide below. The same AI interview engine behind Koji for Education also powers general user and customer research on the main platform at koji.so.

Related resources

• Turning Course Evaluations into NVAO Accreditation Evidence (Netherlands & Flanders)
• Turning Student Feedback into ESG / ENQA Accreditation Evidence
• Course Evaluation Evidence for AACSB & EQUIS Business-School Accreditation
• Course Evaluation Evidence for the UK TEF and QAA Quality Review
• Course Evaluation Evidence for German Accreditation
• Does Closing the Feedback Loop Actually Matter?

Getting started: if your EEE produces numbers but struggles to show action, that is the exact evidence gap French accreditation panels probe. See how Koji for Education turns student feedback into closed-loop, audit-ready evidence at edu.koji.so.