Working with corrected TCF Canada exercises: how to use a corrigé without fooling yourself

Searching for exercices corrigés is a good instinct. Practice with answers attached is the cheapest feedback available, and there is a lot of it about. The problem is what happens next, because most people use the corrigé as the content: they work through a few questions, get stuck, open the answers, read them carefully, nod, and count that as a study session.

That session taught you almost nothing, and it felt productive, which is the dangerous part. A corrigé is not a lesson. It is a measuring instrument, and it only measures if the attempt was finished and closed before you looked.

Why reading answers feels like learning

Recognition is much easier than recall. When you read a worked answer, every step makes sense as it goes past, and your brain records the ease of following it as evidence that you could have produced it. You could not. The judgement I would have got that is made by the same person who did not get it, and it is always generous.

This is why candidates arrive at a real exam having done hundreds of corrected exercises and still freeze. They have practised recognising correct French. The exam asks them to produce a decision under a clock with nothing to recognise.

If you want to feel the difference rather than take my word for it, our exam quiz game is built to make the point against you. Ten questions, twenty-five seconds each, three lives, and nothing revealed until you have committed. The reasoning appears after every answer, which is what a corrigé should be, in the only order that makes it worth anything. Most of the questions test how the exam works rather than your French, and that is deliberate: those are the marks candidates drop at every level, and they are available to you today.

The rule: close the task before you open the answers

Closed means genuinely finished under exam conditions. Timed. One pass on any audio, no pausing and no replaying. No dictionary, no translation tool, no checking a conjugation halfway through. And an answer written for every single item, including the ones you have no idea about, because a blank tells you nothing while a wrong answer tells you what you believed.

One addition makes the whole method work: mark your guesses as you go, a small dot next to any item you were not confident about. When you open the corrigé, a dotted item you got right is not a mark you own. It is a coin flip that happened to land well, and it belongs on the study list with the ones you got wrong. Candidates who skip this step consistently overestimate their level by a band.

Mark the questions you got right

Most people check the corrigé only for errors. The more useful pass goes through the correct answers and asks, for each one, why the other options are wrong. Not why yours is right. Why the others are not.

If you can say it, you have understood the item design, and understood design transfers to the next paper. If you cannot, you probably chose by feel, and feel is exactly what deserts people at minute forty of a real test. In multiple choice sections this single habit does more than any amount of extra material, because these papers reuse the same handful of distractor mechanisms: the verbatim repeat, the true but irrelevant statement, the right fact attached to the wrong person or time, the reversed negation.

What a corrigé cannot tell you

It tells you which items you lost. It cannot tell you why, and why is the only thing you can act on. Supply it yourself, immediately, while you can still remember what you were thinking. Five causes cover nearly everything.

• I did not know a word that the item depended on.
• I misread the stem, or answered a question that was not asked.
• I ran out of time and never really attempted it.
• I heard or read all the words and lost the relation between them, typically a negation, a restriction like ne... que, or who was doing what to whom.
• I had it right and talked myself out of it.

Log the cause rather than the item. Twenty logged causes give you a distribution, and the distribution is a study plan. If most of your losses are time, more vocabulary will not help you. If most of them are the fifth cause, nothing on your list is a knowledge problem at all.

Redo the task, do not reread it

A corrected exercise you have read is not finished, it is loaded. Put it aside for a week and sit it again from a clean copy, timed, with the answers out of reach. If your second attempt is much better, be careful: some of that is the skill and some of it is recall of the answers, and you can tell the two apart by whether the same error type turns up in a set you have never seen.

For listening material, the best second use is not redoing the questions at all. Transcribe thirty to sixty seconds of the audio word for word, then check it against the transcript. The corrigé told you that you missed the item. The transcription tells you which sound boundary you missed, which is the thing you can practise.

Where corrigés stop working entirely

For listening and reading, a corrigé is a real answer key: there is one correct option and you either chose it or you did not. For expression écrite and expression orale there is no such thing. A model answer is one acceptable performance among thousands, and comparing your text to it tells you that yours is different without telling you whether it is worse, or in which criterion, or by how much.

This is the honest limit of self study with corrected exercises. The productive skills are marked against criteria such as range, coherence, accuracy and task fulfilment, and you cannot apply those criteria to your own work, because the meaning you intended is sitting in your head making the text look clearer than it is. That gap is what marked practice on the tiers set out on our pricing page is for.

A week that uses corrigés properly

Two full timed sets, marked to the rules above, with the causes logged. One redo of an older set from a clean copy. One transcription session on audio you failed. One piece of writing or speaking that no answer key can judge, sent to someone who can. Keep the sets regular with the study planner and keep them honest with the exam timer.

Then check the numbers against what your file actually needs. Each ability converts to a Canadian level separately and the weakest one decides, which is why the target is quoted as NCLC 7 in all four rather than on average. Bands and conversions are laid out on the TCF Canada score chart, and the section by section overview is on the TCF Canada page. Confirm current specifications with France Éducation International before you book, and note that we are an independent preparation service with no affiliation to them.