Do Gymnasium preparation courses work? What the studies show

Almost every answer to this question on the Swiss internet comes from somebody who sells a course. We sell no courses and no tutoring. We therefore have no interest in this answer being yes, and none in it being no either.

The honest answer is less comfortable than both: the available studies contradict each other, and the contradiction is the most interesting part of the story.

What is being prepared for here

For readers who have arrived in Switzerland recently, one paragraph of context, because it explains why this market exists at all.

The Gymnasium is the academic branch of upper secondary school and it leads to the Matura, the certificate that gives direct access to the Swiss universities. There is no separate university entrance exam afterwards and no application competition between candidates. The selection happens earlier instead, at the entrance to the Gymnasium: at about twelve for the Langgymnasium, at around fifteen for the Kurzgymnasium. The exam at that door is what everyone calls the Gymiprüfung.

That single morning is what parents are buying courses for. It is also worth saying plainly that it is not the only route to a degree. The apprenticeship, combined with the Berufsmaturität and if necessary the Passerelle, reaches a university too, and it is the route most young people in Switzerland take. Keep that in view while you read the prices further down, because it changes what a course is actually buying.

One more piece of context. Education in Switzerland is a cantonal matter, so whether there is an exam to prepare for at all depends on where you live. In cantons that decide on school marks over a whole semester, this entire market barely exists.

The starting point: 90 per cent

In a survey of around 3‘000 young people in Zurich, 90 per cent said they had attended at least one preparation course. That figure is the key to the whole subject, for a reason that is rarely spoken out loud.

When nine out of ten candidates attend a course, the course is no longer a measure, it is the normal state. A course then gives no advantage over the field, at best it prevents a disadvantage. At the same time the pass rate is determined by the procedure and not by the preparation: when places are limited, the rate cannot rise, however well everybody prepares.

That is not a criticism of the courses. It is the description of a race in which everybody runs faster and the gaps stay the same.

Two sources, two answers

Two statements sit side by side in our fact base that appear at first glance to rule each other out. A study from 2017 finds an effect of preparation courses. The NZZ, by contrast, reported that courses barely increase the chances of success.

We mark that openly: we have not inspected the primary sources of these two positions ourselves for this article. Both count as citable in the debate, and both are cited regularly, depending on which side is arguing. Providers quote the first, sceptics the second.

The usual way of handling that is to pick a side and leave the other one out. The better way is to explain why two serious studies of the same subject can reach different results without either of them having miscalculated.

Why the studies contradict each other

The reason is called self-selection, and it is the central methodological problem of this whole question.

A family that books a preparation course voluntarily already differs, before the first day of the course, from a family that does not. That family understood the procedure early. It has the registration deadlines in view. It can raise several thousand francs. The child usually brings a level of prior attainment that makes the effort look worthwhile in the first place, and often a motivation of its own for the goal.

If you now compare the course group with the non-course group, you measure all of those differences too. The course gets the credit for what partly belongs to the home, to prior attainment and to motivation. Depending on how strongly a study controls for those prior differences, the result comes out differently. That is exactly where the contradiction comes from.

The comparison nobody can make

The course effect could only be measured cleanly like this: take a large group of registered children, split them at random into two halves, send one half to the course and not the other, then compare the pass rates.

Nobody will run that experiment. It would mean withholding from a randomly chosen half a preparation that their parents believe helps. No canton and no provider would take responsibility for that, and it would be hard to justify.

From that follows a statement you will not find in any course brochure: the question of whether preparation courses work cannot be answered conclusively with the data that exist, and it will not be for the foreseeable future. Anyone who quotes you a clear figure is quoting a correlation and selling it as an effect.

What a course costs and what is in it

The range of CHF 1‘980 to 3‘340 is the usual one on the Swiss market for the longer course programmes. What you actually receive within that range differs a great deal: group size, number of mock exams, the qualification of the course leader, and whether there is an individual assessment of where the child stands at the start.

The most important item in a good course is rarely the material. It is getting used to the format, the control of time, and the feedback on marked work.

What you can do yourself for less money

The canton of Schwyz publishes the syllabuses for German, mathematics, French and English together with one specimen paper and its answers per subject. The canton of St. Gallen provides specimen questions, marking schemes and criteria grids for the written composition. The canton of Bern publishes questions from earlier entrance exams.

The sentence the canton of Schwyz puts next to this deserves to be quoted directly: pupils who have mastered this material should be able to pass the exam.

A practical note for families whose German is still developing. All of this material is published in the language of the canton, and the exam itself is written in that language, so working through the specimen papers is worth as much for the language as it is for the subject. If reading the syllabus is a hurdle, the class teacher is the cheaper interpreter, not a course provider.

The cost question is a question of access

If a widespread part of exam preparation costs between CHF 1‘980 and 3‘340, then access to that part depends on a family budget. That is a fact of education policy and not a side issue, particularly in a procedure that helps decide access to higher education.

The most effective counter-measure already exists: the syllabuses, specimen questions and marking schemes of the cantons are public and free. They cover the most important part of the content of what a course offers. What they do not replace is the supervision and the regular feedback.

For families who cannot or do not want to raise several thousand francs, the biggest lever is therefore not buying a course but structure: a fixed weekly slot, real exam conditions, and work that is marked honestly.

What even a perfect study would not answer

Suppose the clean study with random allocation existed, and suppose it found an effect. Even then the question you are asking as a parent would not be answered.

A study delivers an average effect across many children. It says nothing about whether a course works for your child. An average can arise from one group in which a course achieves a great deal and a second in which it achieves nothing at all, and both appear in the mean as a moderate effect.

The more relevant question in practice is a different one: is there a concrete, nameable gap, and is a group course the right instrument for it. A child who is not secure with fractions needs practice at fractions, not two semesters of all round supervision. A child who knows the material but falls apart under time pressure needs exam simulations. Those two cost very different amounts.

The timing counts more than the price

The most expensive mistakes in this area rarely have to do with the price tag. They have to do with timing.

A course that starts four weeks before the exam can no longer build knowledge. It can train the format, sharpen time management and give confidence, and that is not nothing. But it is often sold as preparation in content.

A course that starts two years before the date, on the other hand, ties up free time, money and attention over a very long period for a goal that is still a long way off for an eleven year old. Between those two extremes lies what the cantons themselves provide: a published syllabus that shows what is missing, and specimen papers that let you check it. With that information it is easier to decide whether and when a course makes sense.

How to recognise a usable course

If you do decide on a course, three questions are more useful than any success statistic in the brochure.

First: is there an assessment of where the child stands at the beginning, and what follows from it? A course that treats all participants the same is a textbook with a seat. Second: how many mock exams are written under real time conditions and returned individually marked? That is the part that is hardest to reproduce at home. Third: how does the provider handle the question of effectiveness? A provider who quotes a pass rate without mentioning self-selection has either not understood the question or is not answering it honestly.

What we offer instead

Our contribution to this question is deliberately small. We offer a free reference test across five cognitive areas that takes 25 minutes. It does not tell you whether your child will pass the Gymiprüfung, and it does not replace a cantonal procedure.

What it can do is put a question before the decision about a course that otherwise often comes after it: where are the relative strengths, and where is the weakness that would actually have to be worked on. That information costs nothing, and it is the difference between a course that closes a gap and a course that spends a year repeating what is already secure.

If you are not sure yet whether your canton examines at all, the overview by canton answers that first.