Why Your IB Maths Grade Is Stuck Despite More Practice
By Renzo Diaz·4 min read
"I understood it when the teacher showed me" and "I can solve a new question alone" are different achievements — and a stuck grade despite real practice usually means the gap is in the second one, not the first. Find the first decision in a student's own working that actually failed, then test whether they can make that same kind of decision independently on a different question. More hours of watching worked solutions rarely closes that specific gap.
A parent watching a stuck grade from the outside usually sees the symptom — hours of revision producing the same result — without a way to see the actual cause. The cause is rarely a lack of effort or a gap in content knowledge alone. More often it's a single, specific point in the student's own reasoning where a decision goes unsupported, repeated across different topics in a way that looks like "not enough practice" from the outside.
Recognising a worked solution and producing the method yourself, unprompted, are different skills — and a stuck grade usually means only the first one is solid.
1. What "more practice" is actually missing
The IB's own framing of mathematics assessment centres on non-routine problem solving and mathematical communication, not on reproducing familiar steps. Its published specimen papers are explicit that a correct final answer without supporting working does not necessarily earn full marks — the reasoning has to be visible and defensible, not just implied by a right number.
A student can spend real hours watching worked solutions and still lose marks on an actual paper, because watching a solution and producing one under exam conditions draw on different skills. The useful response to a stuck grade isn't more volume of the same kind of practice — it's finding the first point in the student's own work where a decision actually failed, then testing whether they can make that same decision unaided, on a different question.
2. Four causes that all look like "needs more practice"
A blank page, a wrong final line, a correct number with few marks, and a good score on familiar exercises but a weak one on real papers can all look like the same problem from the outside — but they usually have different causes, and treating them the same way wastes effort in the wrong direction.
| What you see on the paper | Possible cause | A useful next test |
|---|---|---|
| Blank until someone gives a hint | Cannot select a starting method | Ask what feature of the question should suggest each candidate method |
| Right method, wrong last line | Algebra, calculator input, rounding, or units | Retry a new question with the same operation, then check the intermediate step |
| Correct number, few marks | Working or explanation is missing | Hide the final answer and see whether the written argument still makes sense on its own |
| Strong on familiar exercises, weak on real papers | Recognition tied to surface cues, not the underlying idea | Present the same idea in a new context without naming the chapter it came from These are hypotheses worth testing, not a diagnosis available from a grade alone, and a student may genuinely have more than one of them at once. Asking them to talk through a solution before showing the markscheme is where the actual information is — the moment a justified step turns into a guess is the real starting point for instruction. |
3. A 30-minute diagnostic a family can run directly
The process doesn't require a tutor to run once — it requires a recent, independently completed problem with a markscheme that actually matches the student's syllabus and paper. Have the student annotate their own solution: what did I know at this point, and why did I choose this operation? Then mark the first unsupported transition specifically, even if later lines also contain errors — that first break is the actual diagnostic signal, not the total error count.
Explain or rebuild only that specific transition, with one worked example if genuinely needed, then give a different question requiring the same underlying decision and leave the student to choose the approach entirely unaided. If that fifth step fails, they've recognised a demonstration without yet owning the method themselves. If it succeeds, retest after a few days rather than immediately — a correct answer right after coaching is weaker evidence than the same correct answer showing up independently, later.
A correct answer right after coaching is weaker evidence than the same correct answer, unprompted, a few days later.
4. Using AI without confusing recognition with understanding
AI tools can display a full solution instantly, which is exactly why they're easy to misuse for this specific gap — they're excellent for comparing alternative approaches or interrogating a step, and much less useful the moment they replace a student's own attempt at the reasoning. A workable habit is to use the tool, then close it and work a genuinely fresh problem alone. A useful prompt to give the tool itself: don't provide the next step, ask which assumption licenses the method already chosen — treating its answer as something to verify against the syllabus and markscheme, not something to trust outright.
The goal isn't banning assistance. It's making sure the student's own reasoning is still available once the assistance is gone. For a parent checking homework, "could you explain why this method fits" is a far more useful question than "did you get the same final number."
5. Track the change that actually matters
For each recurring error, the useful thing to record is two-fold: whether the student made the correct choice on a genuinely new problem, and whether the written work showed enough evidence to actually earn the marks for it. A few dated, specific examples tracked this way are more informative than any claim that a fixed number of tutoring hours moves a grade by a fixed amount — a claim no legitimate source can actually support for an individual student, because the underlying causes differ from one student to the next.
Sources
The source below is the IB's own current curriculum documentation, checked directly in a browser.