Where IB Mathematics Students Actually Lose Marks — And Why Knowing the Maths Isn't Enough

By Renzo Diaz··9 min read

Most mark loss in IB Mathematics has little to do with not understanding the material. It comes from not demonstrating that understanding in the specific form the mark scheme and IA criteria require — visible method, correct command-term compliance, and a visible trail of mathematical decision-making the examiner can actually see and award.

A student finishes a Paper 2, checks the answers, and walks out confident. They knew the mathematics. They used their GDC correctly. They answered every question. Then the results arrive, and the grade is lower than expected. The first reaction is usually "what did I get wrong?" Sometimes that is the wrong question. The more useful one is: "where did the marks go?"

There is a structural feature of IB Mathematics that is easy to underestimate: knowing how to solve a problem is only part of demonstrating achievement. A student can grasp the underlying mathematics completely and still fail to demonstrate enough of that understanding for the marks to be awarded. This article is not a list of generic study tips. It examines where marks disappear in IB Mathematics — Analysis and Approaches (AA) and Applications and Interpretation (AI) alike — in the Internal Assessment, the written papers, command terms, GDC use, and precision.

The examiner can only award the evidence that appears in your work.

1. The Internal Assessment is not just another maths assignment

The mathematical exploration is a substantial part of both AA and AI, weighted at 20% of the final grade at both SL and HL. The IB describes it as a short report written by the student based on a topic chosen by them, with an emphasis on mathematical communication — accompanying commentary, good mathematical writing, and thoughtful reflection.

Your IA is not simply an opportunity to demonstrate that you can perform calculations. It is assessed on five separate criteria — Presentation, Mathematical Communication, Personal Engagement, Reflection, and Use of Mathematics — and each rewards a different capability. A student who performs well in timed exams can still score poorly here, not because they lack mathematical ability, but because the IA tests something exams do not: whether you can select a focused question, communicate your thinking clearly, take ownership of the work, and reflect honestly on its limitations. A student who has never practised those specific skills starts at a disadvantage that has nothing to do with how well they can solve equations.

Worth knowing if you are earlier in the Diploma: a revised Analysis and Approaches course launches for first teaching in August 2027, with first assessment in May 2029, and its Exploration criteria will be reorganised around a four-stage inquiry process — Problem specification, Abstraction, Computation, and Interpretation — rather than the five criteria above. If you are already partway through the current course, none of that changes what governs your IA. The five criteria above are what is actually being marked.

2. A mathematically correct exploration can still be a weak one

Imagine two students investigating population growth. Both use correct mathematics. Both produce clean graphs. Both obtain correct numerical results. Student A spends the report reproducing a standard model exactly as it appears in textbooks and online examples. Student B defines a specific problem, states assumptions, chooses a model, explains why that model is appropriate, examines its limitations, and refines the approach when the results do not match expectations.

They have not demonstrated the same thing. The IA's criteria are designed to distinguish those two processes, rewarding a student's own mathematical decision-making rather than just correct output. So the better question to ask about your own exploration is not "is my mathematics correct?" It is: "can the examiner see my mathematical decision-making?" That is a much higher standard, and it is the one your report is actually held to.

Can the examiner see my mathematical decision-making?

3. Your choices need to be visible

A strong exploration leaves a trail. Why this model? Why this dataset? Why this assumption? Why this method, and why was an earlier approach rejected? Those decisions are not decorative commentary — they are the actual evidence of personal engagement and reflection that the criteria are looking for.

That changes how you should approach the writing. Do not ask "how much mathematics can I fit into this topic?" Ask "what mathematical problem am I actually investigating, and can someone else see why I made the decisions I made?" A report that only shows the finished calculation, with none of the reasoning that led there, is showing the examiner the wrong thing.

4. The exam is not a multiple-choice test with longer working

In written papers, one of the easiest misconceptions is that the final numerical answer is what matters most. The IB's own instructions to candidates on Mathematics exam papers say otherwise, directly: "full marks are not necessarily awarded for a correct answer with no working. Answers must be supported by working and/or explanations." The same instructions specifically state that "solutions found from a graphic display calculator should be supported by suitable working."

That is a direct statement that visible reasoning is part of what is being assessed, not a formality layered on top of it. It also explains something students often find frustrating: you can obtain the correct number and still not receive full marks. That is not the examiner taking marks away. It is that the work does not contain the evidence the mark scheme requires.

The visible reasoning is part of what is being assessed — not a formality layered on top of it.

5. Correct answer, undemonstrated method

Suppose a question requires a particular result. You type the expression into the GDC, get the correct answer, and write down the number. The examiner cannot award method marks for a method that is not shown, so this often scores less than a student expects.

The reverse can also happen. A student makes an arithmetic error but clearly demonstrates a correct method — and the same instructions to candidates state that "where an answer is incorrect, some marks may be given for a correct method, provided this is shown by written working." That is why checking only the final line of a past-paper answer is such a weak habit. When you review your own work, ask two separate questions: did I get the answer, and did I show enough of the method for the examiner to actually award it? Those are not the same question, and most students only ever ask the first one.

6. Command terms are not decoration

IB Mathematics uses a defined glossary of command terms, and they are not interchangeable ways of saying "do the question." The wording of a question tells you what kind of evidence the examiner is asking you to produce — so a useful habit is to read and underline the command term before you touch the mathematics.

Layer 1 — "Show that"

Obtain the required result — possibly using information given — without the formality of proof. The question is giving you the target; your job is to demonstrate the route to it, not just confirm it works.

Layer 2 — "Prove"

Use a sequence of logical steps to obtain the required result in a formal way — a higher bar than "show that."

Layer 3 — "Justify"

Give valid reasons or evidence to support an answer or conclusion — reasoning beyond simply stating the final value.

Layer 4 — "Verify"

Provide evidence that validates a result — closer to checking than to deriving.

Layer 5 — "Hence"

Use the preceding work to obtain the required result, rather than starting the method over from scratch.

Layer 6 — "Find" vs. "Determine"

"Find" asks for an answer with the relevant working shown. "Determine" specifically means there is only one possible answer to obtain.

7. "Show that" is where working backwards becomes dangerous

This deserves its own section, because it catches students repeatedly. Suppose the question tells you that a quantity should equal a particular expression. You substitute that expression into the preceding formula, and it works. You have verified the answer — but "show that" specifically means obtaining the required result, not merely checking that it holds. Using the given answer as a step inside your own derivation treats the exercise as a verification problem, when the command term is asking for the opposite.

A provided answer is a target, not a shortcut.

8. The GDC can give you the number and still leave you without enough evidence

The GDC is enormously useful, and also easy to misuse. The issue is not that using the calculator is somehow discouraged — the instructions to candidates explicitly expect GDC-based solutions to be supported by suitable working. The issue is what you submit as evidence of your reasoning.

This becomes especially visible with calculator output: raw E-notation (writing 3.2E-4 instead of 3.2 × 10⁻⁴), unlabelled scientific notation, and long decimals copied digit-for-digit from the display all read as transcription rather than understanding, even when the underlying value is correct. Copying the screen is not the same as communicating the mathematics. The habit to build is simple: use the calculator for computation, and mathematical notation for communication. A result should usually connect back to the question too — a bare "14.7" communicates less than "the predicted velocity is 14.7 m/s," particularly in Applications and Interpretation, where the mathematics is explicitly tied to a real-world context.

9. Rounding, exact values, and calculator settings

IB Mathematics exams specify that, unless otherwise indicated, answers should be given exactly or correct to three significant figures. That single rule generates a disproportionate share of mark losses, for two separate reasons. Rounding intermediate values compounds error through a calculation chain, so the safer habit is to store full precision in the calculator and round only the final reported value. And converting an exact quantity into a decimal when the question asks for an exact value — writing 2.646 instead of √7 — is a specific, avoidable error, since the notation itself communicates mathematical information that the decimal throws away.

A related, purely mechanical error is GDC mode confusion — degrees versus radians. The IB does not set a default mode; the question indicates which is expected, and entering a value in the wrong one produces a numerically plausible but wrong answer. That is not a mathematical error in the usual sense. It is a tool-handling error, and the mark scheme does not forgive it regardless of how sound the reasoning behind it was.

Technology does not remove the need for mathematical control. It increases it.

10. What this means for how you prepare

The patterns above share a common thread: they are not about what you know, but about how you demonstrate what you know within the specific constraints of the IB assessment system. Practising past papers by checking only whether the final answer matches the mark scheme tests one part of what the exam rewards. The other part — communication, notation, command-term compliance, and a visible method — needs separate, explicit practice.

A more useful way to review a past paper is to audit the evidence, not just the answer. For each question: did you identify exactly what the command term required? Did you show enough working for the method marks, independent of whether the final number was right? If you used the GDC, did you communicate what it calculated and why it mattered? Did you preserve precision until the final value, and give that value in the required form? For the IA, the same shift applies — do not ask only whether the mathematics is correct; ask whether the examiner can actually see the inquiry behind it.

None of this is solved simply by learning another chapter of mathematics. So when you lose a mark, the more useful question usually is not "what did I get wrong?" It is: "what evidence was missing?"

You do not receive marks for what you know but fail to demonstrate.

Sources

Sources 1, 2, and 3 below are primary — official IB documentation or its direct output. Source 4 is a secondary source on command terms and notation in practice. The six command-term definitions quoted in Section 6 come from the Mathematics: Analysis and Approaches subject guide's official glossary of command terms — a school- and teacher-facing document without a stable public IB web address, checked against multiple independent copies for accuracy before being quoted here. Everywhere else in this article is based on documented patterns observed across IB Mathematics students.