It’s easy, when revising, to keep returning to what already feels comfortable. A topic that looks familiar on the page gets mistaken for a topic that is mastered, and the harder material waits. In IB Mathematics: analysis and approaches, that habit gets expensive. The subject guide says exams award marks for method, accuracy, reasoning and interpretation, and the curriculum updates emphasise problem-solving, inquiry and applying techniques in context. Recalling a procedure is rarely enough when a question arrives in an unfamiliar setting. A repeatable revision cycle prepares you for that better than a string of disconnected study sessions.

That cycle draws on five practices. You diagnose where you really stand, space your practice across weeks, and mix topics rather than drilling them one at a time. You also log errors by type and rehearse under timed conditions.

Start with an Honest Diagnosis

Start by finding out what you can actually do under exam-like conditions. Attempt a small set of past-paper questions without notes, so you can see which kinds of problems you can complete and which stall you. Because the IB awards marks for method and reasoning as well as answers, a rough feel for an answer is not the same as being able to collect those marks.

The Weekly IB Math Revision Loop turns that diagnosis into a routine. A weekly rhythm keeps topics in play instead of letting them drift until a last-minute cram.

1. Diagnose (30–45 minutes, once a week): Attempt 3–5 short past-paper questions cold, with no notes. Include one from a current weak area and one from an older topic you want to keep fresh. Tag each attempt: (A) couldn’t start, (B) partial method but stuck, (C) solved but slow or fragile, (D) clean.
2. Choose this week’s targets (5 minutes): Pick two topics. The first, your repair topic, comes from (A) or (B): something you’re rebuilding because you couldn’t start or got stuck. The second, your stabilise-and-speed topic, comes from (C): something you can already do but want to make quicker and more secure. If more than two topics sit in (A), take the ones that recur most in recent attempts rather than fixing everything at once.
3. Guided rebuild (2 sessions): For the repair topic, work through scaffolded questions until you can earn method marks without prompts. End each session with one unseen question done from memory instead of more re-reading.
4. Interleave to train method recognition (2 sessions): Build a mixed set with 50–70% of questions from this week’s two targets and 30–50% from older topics. After each question, write one line: “What feature told me what method or formula to use?”
5. Timed rehearsal (1 session): Run a timed mini-simulation, either a full section of a past paper or a fixed number of mixed questions under a timer. Use the formula booklet as you would in the exam, under the same time pressure you’ll face on the day.
6. Close the loop with your error log (10–15 minutes): Log every lost mark by error type (conceptual, terminology, or computational slip) and write the next action. Use the log to choose next week’s repair topic before you choose anything else.

Notice how often that loop sends you back to a topic you’ve already met. That’s deliberate, and the research on spaced practice and retrieval explains why those returns carry so much of the value.

Build Your Revision Across Weeks, Not One Topic at a Time

Finishing one topic block and moving on feels productive, but it is a form of massed practice. A meta-analytic review in Educational Psychology Review covered 27 mathematics studies. It found that spaced practice produced a consistent small-to-medium advantage over massed practice for retention and test performance. Retrieval practice also tended to improve outcomes, although its effects varied across implementations. So a revision schedule that returns to topics across several weeks tends to work better than doing all the practice for each topic in a single block.

Leaving a topic alone just as it starts to feel secure sounds like negligence. On this evidence, it’s closer to a method.

A simple rotation puts this into practice. Introduce a topic, then come back to it after a gap with fresh problems. Later, revisit it again inside a mixed set. At every return, try problems from memory before you open your notes. Self-testing then becomes a built-in part of the cycle, and the gaps it exposes tell you what to re-read.

Mix Topics in Practice and Log Every Error by Type

A topic that survives a return visit still has to survive company. Once you’ve re-read what the gaps showed you, move to mixed past-paper questions and eventually full papers. A meta-analysis of ten learning techniques in Frontiers in Education found that interleaving helps students distinguish among problem types and select appropriate methods, even when practice feels harder and short-term performance looks worse than with blocked drilling. IB exams expect you to choose suitable techniques without being told the topic label, so mixed-topic sets teach you to recognise the features that signal each method.

Mixed practice also produces more mistakes to learn from, and an error log makes each one useful. The study Secondary School Mathematics Teachers’ Accuracy at Predicting Student Errors, published in School Science and Mathematics, categorises common secondary mathematics errors as conceptual misconceptions, vocabulary or terminology misunderstandings and computational slips. It also finds that teachers are only moderately accurate at predicting which specific errors students will make. So analyse the mistakes you actually make instead of trying to predict them.

Each of the three error types calls for a different kind of corrective work. A concept gap needs rebuilding through guided questions or structured practice. A terminology misunderstanding needs deliberate work on the words themselves, including IB command terms. A computational slip needs a checking habit. The IB does not define command terms in the exam, so that vocabulary has to be learned during revision. Calling every lost mark careless is comforting, but it hides three different problems. A log pays off only when the patterns in it trigger specific next actions.

• Logging, right after each practice set (about five minutes): record the paper and question ID, and note which kind of mark you lost. Tag the error as concept, terminology or command term, or computational slip, then write one next action in a single sentence.
• Review, twice weekly for 10–15 minutes: scan the last 10–20 entries and circle repeats, whether in the same error type, command term or topic.
• Three or more concept errors on one skill: schedule a guided rebuild session before doing more timed work on it.
• Repeated terminology or command-term errors: add a 10-minute command-term drill before your next paper session, then redo those questions with the command term explicitly satisfied.
• Repeated computational slips: choose one mandatory check step, such as signs, units, rounding or graph window, and reattempt two questions under time pressure to test the habit.
• One topic appearing across several error types: make it a priority in next week’s loop, because it is unstable rather than merely careless.

Even so, a log can only record what went wrong. It can’t tell you whether a fix still holds when the clock is running.

Run Timed Simulations to Practise Showing Your Working

Timed simulations are where you practise showing your working and reasoning under real time pressure, rather than chasing only the final number.

In the exam, students receive a clean formula booklet and are expected to know it thoroughly beforehand. A condensed IB math formula booklet from a revision site can help you learn where formulas sit, but it is only a study aid and not the official booklet permitted in exams. It stays at home on the day. Rehearsing with a booklet you can’t use on the day is rehearsing with a prop, so every simulation should use the official one.

Pacing is the other half. The curriculum updates report an assessment change that reduces the combined marks for Papers 1 and 2 to a total of 75, so each mark is worth more and time management matters more. Run each simulation in one sitting with only allowed tools. Set two checkpoints, one at the midpoint and one in the last 10–15 minutes, and use them to decide whether to skip a question and bank marks elsewhere. Booklet lookups need a rule too. Name what you need before you open it and return to the question within 20–30 seconds. If you can’t, mark the question and move on.

The debrief is where the simulation pays off. Self-mark with the markscheme and look for any method, reasoning or interpretation marks you lost even when the final answer was close. Log each loss with a next action. Then build the next session around your two or three costliest mistakes: reattempt them briefly, and add one fresh mixed set to test whether the fixes transfer.

Begin the Cycle with a Cold Diagnostic

The payoff of this cycle is that every lost mark tells you what to do next. Spacing and mixing show whether a skill lasts and whether you can pick the right method. Timed practice with the official booklet shows what holds up under exam conditions. The sequence matters because the error log chooses next week’s repair topic, so each week’s work is decided by the previous week’s mistakes.

Your first move is simple. Attempt a few short past-paper questions cold, and tag each one by how far you got. A topic that looks familiar on the page will either pass that test or fail it, and it’s better to find out now than on the day.