How to Write CMA Case-Based Questions: The Scoring Rules and a Worked Approach

Of everything in the CMA exam, the case section is the part candidates understand least — and it shows in their scores. Students will tell you the MCQ rules down to the minute, then describe CMA case-based questions with a shrug: 'two essays, an hour, write something sensible.' That shrug is expensive. The case section carries 25% of your score, sits behind a gate you must clear first, and rewards a specific kind of performance — methodical, structured, calculation-first — that most candidates never practise deliberately. I have gone through IMA's published guidance on how this section works and how it is graded. Here are the scoring rules as they actually operate, a step-by-step framework for any scenario, and one fully worked example — an invented teaching case, clearly labelled — showing the same answer written weakly, then written the way the scoring rewards.

What the Case Section Actually Is in 2026

Each CMA part ends with a one-hour section built around two business scenarios: a short case — a company, some data, a problem — followed by several tasks asking you to calculate, analyse, explain or recommend. The section is worth 25% of your score; the 100 MCQs in the preceding three hours carry the other 75%. Combined performance converts to a scaled score out of 500, and 360 passes.

Now the change every 2026 candidate must plan around. For decades this section meant free-text essays. IMA is replacing them with Case-Based Questions (CBQs) on a phased timeline through 2026, and where you sit in that timeline decides what you will actually face:

The exception: China, Japan and Taiwan continue with essays for now, with CBQs phasing in during 2027. For Knowledge Valley's students — sitting English-language exams in the Gulf or Europe — the practical conclusion is simple: if your window is September/October 2026 or later, you are a CBQ candidate. Prepare for CBQs.

What does a CBQ actually look like? One short business case of roughly 250 words, sometimes with exhibits such as a data table, followed by around six to seven structured questions on that case. Instead of writing paragraphs, you respond in set formats: entering a calculated number, dragging items into order or categories, selecting from a list, or choosing multiple correct options. IMA's stated position is that content, difficulty and the passing standard are unchanged — same syllabus, same 25% weight, same one hour, same 50% MCQ gate. Only the way you demonstrate your knowledge has changed. Does classic 'essay technique' still matter, then? Almost entirely, yes — because the thinking the framework below teaches (read the requirement first, extract the data, work methodically, conclude) is exactly what structured case questions reward. What changes is the output: precise entries instead of prose.

And the gate, because it frames everything: you see this section only if you answered at least 50% of the MCQs correctly. Fail to clear it and the exam ends after three hours. The case section is not a bonus round; it is a quarter of the qualification, locked behind the part of the exam most within your control.

The Scoring Rules: How Your Answers Are Actually Graded

IMA does not publish a mark-by-mark rubric — anyone claiming to have the exact internal scheme is guessing — but the operating rules are well established from IMA's handbook guidance and grading descriptions. Each one has a direct behavioural consequence.

• Human graders, structured key. For the essay format, answers were graded by trained human graders under ICMA (IMA's certification arm), using a structured answer key, with calibration on sample responses and additional review for borderline scripts. A person read your reasoning — which is why reasoning had to be visible. CBQs replace that judgement call with objective scoring of structured responses.
• Partial credit is the whole game. Graders award marks for correct steps, correct method and partially correct work, not only for correct final answers. A calculation set up correctly but finished with an arithmetic slip still earns most of its credit. The mirror image also holds: a correct final number with no working shown earns little, because the grader cannot verify the method.
• Content accuracy and communication, together. Two things are assessed at once: whether your analysis is technically right, and whether you communicated it like a professional — organised, labelled, using correct terminology. A CMA is expected to explain a variance to a manager, not merely compute one. Rambling, unlabelled answers bleed marks even when the underlying knowledge is present.
• Every sub-question carries its own marks. Each scenario bundles several distinct asks, and marks attach to each one separately — answering three of five tasks well beats answering one brilliantly and ignoring the rest. This is explicit in the CBQ design: each question inside a case is scored independently, so a wrong answer on one does not cascade and cost you marks on the next. Graders cannot award marks for the question you wished had been asked.
• Blank is the only guaranteed zero. An attempted answer with sound steps earns something; a blank earns exactly zero, with certainty. There is no penalty for a wrong attempt beyond the marks that step does not earn.

Read those five rules again and a strategy writes itself: structure visibly, show method always, cover every sub-question, communicate cleanly. That is not exam folklore — it is the grading system, read backwards.

The Framework: A Step-by-Step Approach That Works on Any Scenario

Here is the answering framework I teach from that research. It has four moves, and it is deliberately boring — boring is what survives time pressure.

Move 1 — Read the requirements before the case.

Your first act on opening a scenario is to scroll to the questions and read every sub-question carefully. Note exactly what each one demands: a calculation, an explanation, a recommendation, a list of factors. This takes two minutes and transforms how you read the case, because you now read with a shopping list — hunting the data each question needs instead of absorbing the story passively.

Move 2 — Extract and organise the data.

As you read the case, pull the relevant numbers and facts out beside each sub-question: budgeted figures, actual figures, rates, capacities, constraints. Note what is given — and what is not given and must be derived. Many scenario questions hide a two-step calculation: a rate must be built from data before the headline calculation can start. Listing the inputs first exposes those hidden steps while you are calm, not mid-calculation.

Move 3 — Answer under structure: heading, calculation, sentence.

For each sub-question, use the same skeleton. A heading that restates the task ('1(a) Material price variance'). Then the calculation, line by line, each input labelled, formula visible — not a leap from data to answer. Then a sentence or two interpreting the result: favourable or adverse, why, and what it means. Under CBQs, translate the same skeleton into your scratch work: heading, labelled steps, then the final entry. Method first, answer second — in both formats.

Move 4 — Conclude like a professional, then move on.

If the task asks for a recommendation, give one — clearly, in the first sentence of your conclusion, supported by your numbers. Hedging paragraphs that end without a position are the most common weakness I see in practice answers. Then watch the clock: roughly 30 minutes per scenario is your budget, and a good answer to scenario two is worth more than a perfect answer to scenario one plus a blank page. When time is short, prioritise breadth — a headed skeleton with key steps for every remaining sub-question banks more partial credit than one polished answer.

A Worked Example: One Case, Two Answers

IMPORTANT — PLEASE READ FIRST:

The scenario below is invented by me for teaching purposes. It is not a real CMA exam question, it is not reproduced or adapted from any IMA material, and no real exam question should ever be shared or sought out. It is deliberately small so the difference between the two answers is easy to see.

Illustrative scenario (invented):

Gulf Packaging LLC produces printed cartons. For March, the production budget was 10,000 cartons using 2 kg of board per carton at a standard price of $5.00 per kg. Actual results: 10,500 cartons were produced; 22,050 kg of board were purchased and used at an actual price of $4.80 per kg. Task: (a) calculate the material price variance and the material usage variance; (b) state whether each is favourable or adverse; (c) give one plausible operational explanation for the usage result.

The weak answer (what graders dread):

'The material variance is $17,850 adverse because the company used more material than it should have and production was higher than budget so costs went up. The price was lower which helped a bit but overall materials cost too much this month and management should look into it.' Notice everything wrong here: the two variances are merged into one invented total, no formula is visible, the price effect (actually favourable) is waved at vaguely, no step can earn partial credit, and higher output — which legitimately requires more material — is never separated from genuine waste. If the headline number is wrong, this answer earns close to nothing, because there is nothing else to reward.

The strong answer (same knowledge, grader-friendly structure):

• (a)(i) Material price variance. Price variance = (Standard price − Actual price) × Actual quantity purchased = ($5.00 − $4.80) × 22,050 kg = $0.20 × 22,050 = $4,410 Favourable. The board was bought for less than standard.
• (a)(ii) Material usage variance. Standard quantity for actual output = 10,500 cartons × 2 kg = 21,000 kg. Usage variance = (Standard quantity − Actual quantity used) × Standard price = (21,000 − 22,050) × $5.00 = 1,050 kg × $5.00 = $5,250 Adverse. More board was consumed than the standard allowance for the output achieved.
• (b) Direction. Price variance: Favourable. Usage variance: Adverse. Net material cost variance: $840 Adverse ($5,250 A − $4,410 F).
• (c) Explanation. Production used about 5% more board per carton than standard. A plausible operational cause is a rise in cutting waste or spoilage — for example, a misaligned cutting machine or a batch of lower-grade board producing more rejects — which purchasing savings could not offset. This should be investigated with production, not blamed on the purchasing department, whose variance is favourable.

Compare the two honestly. The knowledge required is identical — two formulas and one sensible comment. The strong answer wins because every step is visible and labelled, each sub-question is answered where it is expected, units and directions are stated, and the conclusion interprets rather than repeats. Had its final arithmetic slipped, it would still have kept most of its marks. That asymmetry — method as insurance — is the economics of this section. In CBQ terms, the same discipline shows up as completing each structured task in sequence rather than hunting for one big answer.

Time, Terminology and the Details That Quietly Decide Marks

A few operational rules tie the framework together. First, terminology: use the syllabus's own words — 'contribution margin', 'relevant cost', 'adverse variance' — because scoring keys are written in IMA's language. Second, calculations: show units and label each line, so your logic can be followed without reconstruction; in a CBQ numerical entry, that discipline becomes doing the full working on your scratch area before typing the number. Third, respect the task format exactly — a stated number of points, a required order, a table — structure is part of the instruction. Fourth, your hour: spend the first two minutes of each scenario on Moves 1 and 2, and protect the last minutes for checking entries and conclusions.

Finally, preparation: none of this is learnable by reading about it — including this article. Case skill is built by completing timed scenarios and having them checked against the rules above: requirement read first? Every task attempted? Workings done before entries? One timed scenario a week through the second half of your preparation is the minimum dose I would defend. Candidates who treated the case section as an MCQ afterthought are the same candidates surprised by a 340. It is trained in exactly one way: repetitions, timed, marked, corrected.

Frequently Asked Questions

How are CMA case-based questions scored?

The section is worth 25% of your part score, combined with the MCQ section's 75% and converted to the 0–500 scaled score, with 360 required to pass. Each question inside a CBQ case is scored independently — one wrong answer does not reduce marks on the others. On timing: essays applied in January/February 2026, candidates chose in May/June 2026, and from September/October 2026 CBQs are the only format on English-language exams (China, Japan and Taiwan excepted).

Who grades the CMA essay/case section?

Essays were graded by trained human graders working under ICMA to a structured answer key, with calibration on sample scripts and extra review for borderline cases — one reason results took roughly six weeks. CBQs use structured response formats scored objectively against set criteria, each question independently, which is why IMA expects noticeably faster score release. Either way, partial credit principles reward correct steps. Verify your window's format at imanet.org.

What happens if I do not reach the essay section?

If you answer fewer than 50% of the MCQs correctly, the case section does not open and the part is failed. Candidates are not told their exact MCQ percentage at the time; if the case section never appeared, that is the reason.

How long do I get for the two case scenarios?

One hour in total, after the three-hour MCQ section. Budget roughly 30 minutes per scenario, including two minutes at the start to read the requirements and extract the data. You cannot return to the MCQ section once you have left it.

Do I get marks for showing my calculations?

Under essays, yes — partial credit rewarded visible method, and a correctly set-up calculation with an arithmetic slip kept most of its marks. Under CBQs, each structured question is scored on its own answer, so the equivalent discipline is accuracy per task: do full workings on scratch paper before entering a number, because only the entry is scored.

What does 'communication skills' mean in CMA scoring?

In the essay era it meant answers that were organised and professional: each sub-question under its own heading, correct terminology, results interpreted rather than left as naked numbers, a clear recommendation where asked. Under CBQs the same quality shows up differently — precise entries, correct classifications, complete task coverage. Either way, it is about whether a manager could act on your work.

Are the worked examples in this article real exam questions?

No. The scenario in this article was invented for teaching and is labelled as such. Real CMA exam questions are confidential IMA material; candidates agree not to disclose them, and any resource claiming to sell 'real' questions should be avoided on both ethical and practical grounds.

Final Verdict

Here is the sentence to carry into your exam: the case section does not reward the candidate who knows the most — it rewards the candidate whose knowledge the scoring can see. Structure is visibility. Method is visibility. Completing every task is visibility. Practise the four moves until they are automatic, and the 25% of your score that frightens other candidates becomes the most controllable part of your exam.

This skill is trained fastest with feedback. At Knowledge Valley, case practice is built into the programme, not bolted on: our unique testing and assessment system includes timed scenario work checked against exactly these rules, and mentorship beyond the classroom means your work gets reviewed by instructors with real corporate experience. With Gulf-friendly timings for working students and smart exam-focused notes that teach the framework topic by topic, we prepare you for the section as it is scored — CBQ format included — not as it is rumoured to be. If your mock case work has never been properly reviewed, bring it to us and we will show you where your marks are hiding.

Researched and compiled by Irfan Haider — SEO Research — for Knowledge Valley

Research note: Section mechanics (two scenarios in 1 hour, 25% of the score; 100 MCQs in 3 hours at 75%; the ≥50% MCQ gate; scaled scoring 0–500 with a 360 pass mark) and grading principles (structured answer key, trained human graders for the essay format with calibration and borderline re-review, partial credit for shown method, assessment of content accuracy and communication) are per IMA / ICMA CMA Handbook guidance and IMA-published descriptions of the grading process, cross-checked October 2026. IMA does not publish a per-question scoring rubric, so internal CBQ scoring is described here as principles, not exact schemes. Format note: IMA's essay-to-CBQ transition timeline (Jan–Feb 2026: essays only; May–Jun 2026: candidate choice; Sep–Oct 2026 onwards: CBQs only on English-language exams; China, Japan and Taiwan retain essays with CBQ phase-in during 2027) and CBQ format details (~250-word case, around 6–7 structured questions, independent scoring per question) are per IMA's announcements and help-centre guidance as summarised by Gleim and Becker, cross-checked October 2026 — confirm your format at imanet.org. The worked scenario in this article is an original, invented teaching example, not IMA material. Sources: imanet.org, IMA/ICMA publications — October 2026.