Evaluating the One-to-One Lesson: Course Evaluation in the Conservatoire

Bottom line up front: The defining teaching form of a conservatoire is the one-to-one instrumental or vocal lesson — one teacher, one student, often the same pair for years, behind a closed door. A course-evaluation instrument designed for a 200-seat lecture hall misfires here for three concrete reasons: the sample is tiny (a class of one, or a handful), the power and dependency dynamics are unusually intense, and the relationship is longitudinal rather than term-bounded. Evaluating higher music education well means designing for those features, not pretending they are noise. This is a distinct problem from evaluating the crit or the studio, and it deserves its own answer.

Why the conservatoire is not the lecture hall

In most of higher education, the signature pedagogy is a group form — the lecture, the seminar, the lab. The conservatoire's signature pedagogy is a dyad. As Helena Gaunt's foundational studies of one-to-one tuition documented, instrumental and vocal students experience their lessons as an intense, individualised relationship, while their teachers often work in professional isolation — teaching alone, rarely observed, with little shared vocabulary for what "good" one-to-one teaching even is. That isolation is the crux: the very feature that makes the lesson powerful also makes it opaque to conventional quality assurance.

Now apply a standard end-of-semester Likert survey to that setting and watch it break:

• The n is one. A principal-study teacher may teach three or four students. A "mean satisfaction" over four responses is statistically meaningless, and the uncertainty around small-class evaluation numbers swamps any signal. Worse, with so few respondents, anonymity collapses — a teacher can often infer who said what.
• The dependency is total. Your principal-study teacher shapes your technique, your repertoire, your professional network, your references, sometimes your auditions. That is fertile ground for the mum effect — the documented reluctance to deliver negative feedback to someone who holds power over you. Honest criticism of the one relationship a student most depends on is precisely the feedback a survey is least likely to elicit.
• The clock is wrong. A survey asks about "this module, this term." A conservatoire relationship unfolds over three or four years. Retrospective, term-bounded ratings miss the arc — the plateau in year two, the breakthrough in year three — that is the actual object of interest.

The European quality context: AEC and MusiQuE

Higher music education is not without external quality machinery. The Association Européenne des Conservatoires (AEC) is the sector's European association, and MusiQuE – Music Quality Enhancement is the independent European accreditation and evaluation body created in October 2014 by the AEC together with the European Music Schools Union and Pearle*–Live Performance Europe. MusiQuE conducts programme and institutional reviews across European conservatoires, from Madrid to Antwerp to Malmö, using standards tailored to higher music education.

But — and this is the point — MusiQuE and AEC standards operate at the programme and institution level. They expect an institution to gather student feedback and use it; they do not, and cannot, prescribe how you make the individual one-to-one lesson legible to quality enhancement. That gap is the institution's to fill, and it is where most conservatoires fall back on either a lecture-hall survey that does not fit, or informal corridor knowledge that does not scale and does not document.

But doesn't standardised feedback threaten artistic apprenticeship?

The strongest objection comes from inside the tradition. The master-apprentice relationship, its defenders argue, is not a "service" to be rated. Turning the lesson into a customer-satisfaction transaction risks flattening a subtle artistic relationship into a Net Promoter Score — and we have argued at length that NPS is the wrong instrument for education and that the student-as-consumer frame distorts what evaluation is for. Won't formal course evaluation of the one-to-one lesson do exactly that damage?

It will — if you import the lecture-hall instrument wholesale. But that is an argument against the wrong instrument, not against evidence. The alternative to a satisfaction survey is not corridor gossip; it is better evidence, gathered in a form that respects the pedagogy. The questions worth asking are formative and developmental: Is the repertoire stretching you appropriately? Do you understand why your teacher is asking for a technical change? Do you feel able to raise a disagreement about interpretation? Are you being prepared for a professional life, or only for your teacher's taste? Those are not consumer questions. They are the questions a reflective conservatoire would want answered — and they are almost impossible to surface through a five-point scale, and almost impossible to surface honestly through a channel the teacher controls.

A second, fair objection: with a class of one, can you ever protect anonymity enough to get candour? Honestly, not perfectly — this is a real limitation, and any responsible system should say so. The mitigation is to aggregate at the level of the studio or department, to collect formatively and mid-relationship rather than as a high-stakes verdict, and to route feedback through a neutral instrument rather than the teacher. You reduce the exposure; you do not eliminate it. Claiming otherwise would be dishonest.

Designing questions that fit a dyad, not a lecture

If the lecture-hall survey is the wrong instrument, what does the right set of questions look like? Not "rate your teacher 1–5." The productive questions in a one-to-one setting are developmental and relational, and they work best when a follow-up can chase the interesting answer. A few that a conservatoire could actually use: How has your relationship with your repertoire changed this year — and did your lessons drive that? When your teacher asks for a technical change you do not yet understand, what happens next? Where do you feel you are being prepared for your own musical voice, and where for someone else''s taste? Is the frequency and length of your lessons serving your development, or has it drifted? Each of these invites a story, not a score, and each is diagnostic of something a department can act on — mismatched repertoire, unexplained technical demands, over-dependence on a single aesthetic, timetable erosion. Crucially, they are answerable honestly only if the channel is neutral and the student trusts that a candid answer will not sour the very relationship they depend on. That trust condition is the whole design problem, and it is why the instrument matters as much as the questions.

What Koji offers higher music education

Koji for Education is well-suited to the conservatoire's problem precisely because it is not a survey. It is a conversational instrument, and conversation is how you evaluate a relationship.

• AI-moderated conversational interviews ask the developmental questions and probe — "you mentioned the repertoire feels repetitive; can you say more about what you would want to be playing?" — surfacing the formative detail a scale cannot. The AI moderator is also a neutral party: not the teacher, not a peer the student will see at tomorrow's masterclass, which directly softens the mum effect.
• Thematic analysis at the studio and department level aggregates individual conversations into patterns — is a whole cohort under one pedagogue reporting the same gap? — giving MusiQuE- and AEC-facing quality processes real evidence without exposing any single student. This is the depth-at-scale that focus groups promise but rarely deliver.
• Formative, mid-cycle collection fits the longitudinal relationship: check in during year two, not only at the end, so problems are caught while the relationship can still adapt. This is the formative-over-summative stance the one-to-one lesson demands.
• Six structured question types (open-ended, scale, single- and multiple-choice, ranking, yes/no) let you combine a little structure with a lot of narrative — ranking repertoire priorities, then talking about why.
• GDPR-compliant, EU-appropriate data handling — essential when the respondent pool is so small that re-identification is a live risk.

Be precise about the claim: Koji does not evaluate the artistry of a performance, and it does not replace the expert judgement of a jury or an external examiner. What it does is make the student's experience of a one-to-one pedagogy visible, developmentally and at institution scale, without the survey artefacts that make lecture-hall tools fail in the studio. The conservatoire evaluation problem is closest to the doctoral-supervision problem — another intense, long, dyadic, hard-to-observe relationship — and the same conversational method fits both.

Conservatoires that use this approach for the one-to-one lesson often extend the same AI interview engine — shared with koji.so — to alumni and employer research about professional readiness. If you are trying to make the studio legible to quality enhancement without flattening it, see how Koji for Education approaches it.

Frequently asked questions

Why don't standard course-evaluation surveys work for conservatoire one-to-one lessons?

Three reasons. The sample is tiny — a principal-study teacher may have only three or four students, so averages are meaningless and anonymity collapses. The dependency is intense, triggering the "mum effect" where students withhold honest criticism of a teacher who controls their technique, repertoire and references. And the relationship is longitudinal, spanning years, while surveys ask about a single term. Lecture-hall instruments are simply the wrong tool.

What are AEC and MusiQuE?

The Association Européenne des Conservatoires (AEC) is the European association for higher music education. MusiQuE — Music Quality Enhancement is an independent European accreditation and evaluation body created in October 2014 by the AEC, the European Music Schools Union and Pearle*-Live Performance Europe. Both operate at programme and institutional level; they expect institutions to gather and use student feedback but do not prescribe how to evaluate the individual one-to-one lesson.

Doesn't formally evaluating the one-to-one lesson damage the master-apprentice relationship?

It can — if you import a lecture-hall satisfaction survey wholesale, which reduces a subtle artistic relationship to a rating. The alternative is not gossip but better, formative evidence gathered in a form that fits the pedagogy: developmental questions about repertoire stretch, understanding of technical feedback, and preparation for professional life, routed through a neutral instrument rather than the teacher.

Can you protect student anonymity when a class has only one student?

Not perfectly, and a responsible system should say so. The mitigation is to aggregate feedback at the studio or department level rather than the individual lesson, to collect formatively and mid-relationship rather than as a high-stakes end verdict, and to route it through a neutral channel. This reduces re-identification risk substantially but does not eliminate it.

How does Koji suit higher music education?

Koji is a conversational instrument rather than a survey, which suits evaluating a relationship. Its AI moderator is a neutral party that softens the reluctance to criticise a powerful teacher, it probes formative detail a scale cannot capture, it aggregates individual conversations into studio- and department-level themes for MusiQuE- and AEC-facing quality processes, and it supports mid-cycle formative collection appropriate to a multi-year relationship — all with GDPR-compliant handling suited to very small respondent pools.