Best-of-Five MCQs: 8 Reasons They’re Harder Than They Look (and How to Beat Them)
Exam technique
8 Reasons Best-of-Five MCQs Are Harder Than They Look
Best-of-five questions are the format behind most UK and international postgraduate medical exams sat by doctors in the UAE, from MRCP to MRCS and MRCPCH. On the surface they look like ordinary multiple choice. In practice, they punish candidates who read like they would read a textbook. Here is what actually goes wrong, and the method that fixes it.
The core problem
Best-of-five is not single-best-answer in disguise
In a classic single-best-answer question, four options are wrong and one is right. In a well-written best-of-five item, two, three, or even four options can be defended in isolation. Only one is the best answer given the full clinical picture and, crucially, given exactly what the question stem is asking for.
That distinction matters. “Most likely diagnosis”, “most appropriate initial investigation” and “most appropriate next step in management” are three different questions on the same vignette, and they can have three different correct answers. Candidates who match keywords instead of parsing the question routinely pick a technically correct option that does not answer what was actually asked.
The 8 traps that cost marks
- Answering the first plausible option. The brain locks onto option B because it fits the third line of the stem, and then quietly discounts everything after. Best-of-five vignettes save the pivot detail for the last sentence roughly half the time. If you commit early, you will read the rest of the stem looking for confirmation, not information.
- Confusing diagnosis with management. The vignette screams “pulmonary embolism”, so you pick heparin. But the question asked for the most appropriate investigationand heparin is not an investigation. Empirical treatment might still be the answer if the stem describes haemodynamic collapse, but only because the question specifically asked for the next step. Read the last line first.
- Anchoring on the previous question. You just answered three cardiology questions in a row. The next stem mentions chest pain, and your pattern-matching says “cardiac” before you have read the vitals. Best-of-five papers deliberately cluster and then break topics to catch this bias. Reset between items.
- Ignoring the setting. “A 68-year-old man presents to a rural clinic” and “a 68-year-old man presents to a tertiary centre” change the correct answer, because the available investigations and the appropriate escalation differ. In UAE practice this matches real life: what you do at a primary health centre in Al Ain is not what you do at a tertiary hospital in Abu Dhabi.
- Missing the negative qualifier. “Which is the least likely diagnosis?” and “which is not a recognised feature?” show up more often than candidates remember. Speed-reading turns them into positive questions and guarantees a wrong answer. Circle the qualifier as you read.
- Weighing options against each other instead of against the question. Two options look similar, so you compare them head to head and pick the one that sounds smarter. The correct approach is to compare each remaining option against the exact wording of the stem. The “smarter” option often answers a question that was not asked.
- Overweighting rare conditions. Postgraduate exams do test zebras, but the base rate still matters. If the stem describes textbook features of a common condition and a single atypical feature that could fit something rare, the common answer is usually right unless the atypical feature is unambiguous. Do not out-clever the question.
- Running out of time on early questions. Spending four minutes on question three means guessing question one hundred and eighty. The average time per item in most postgraduate papers is around 90 seconds. Flag, move on, come back. This is a technique skill, not a knowledge skill, and it only develops under timed conditions.
Method
A four-step approach that works under pressure
Once you know the traps, the method almost writes itself. The point is to run the same routine on every question so that time pressure does not force you into shortcuts.
- Read the last line of the stem first. Know what is being asked before you read the vignette.
- Build a working diagnosis or plan from the vignette before you look at the options.
- Eliminate the clearly wrong options first, usually two or three.
- Compare each remaining option against the exact wording of the question, not against each other.
A worked example
A 34-year-old woman presents to the emergency department in Dubai with pleuritic chest pain and shortness of breath that started two hours ago. She returned yesterday from a 9-hour flight from London. She is on the combined oral contraceptive pill. Her observations: heart rate 118, blood pressure 104/68, respiratory rate 24, oxygen saturation 93% on room air, temperature 37.1°C. Chest examination is unremarkable. ECG shows sinus tachycardia. What is the most appropriate next step in management?
Options:
- D-dimer
- CT pulmonary angiogram
- Therapeutic low molecular weight heparin
- Bedside echocardiogram
- Ventilation-perfusion scan
How to work through it. The last line asks for the next step in management, not for the diagnosis and not for the confirmatory investigation. The clinical picture is a high-probability pulmonary embolism with borderline haemodynamic compromise: tachycardia, low-normal blood pressure, hypoxia. Wells score is high.
Eliminate D-dimer first: in a high-probability patient it is not useful, a negative result would not change management. Eliminate ventilation-perfusion scan: it is a reasonable alternative to CTPA in specific situations, but not the immediate next step here. That leaves CTPA, echocardiogram, and empirical anticoagulation.
Echocardiogram is useful if the patient is too unstable to move to the scanner, but she is borderline, not peri-arrest. CTPA is the definitive investigation, but the question is about the next stepand in a high-probability patient with clinical instability, guidelines from bodies including the UK NICE venous thromboembolism guideline recommend starting empirical anticoagulation before imaging if imaging will be delayed. In a busy emergency department that delay is realistic.
The best answer is therapeutic low molecular weight heparin. Notice that CTPA is not wrong, it is just not the best next step. That is the whole shape of best-of-five.
Why practice content is not enough
Technique degrades under time
Reading the last line first is easy when you are relaxed on the sofa. At minute 140 of a 3-hour paper, with 40 questions still to go, the habit collapses unless it has been drilled under real conditions.
Untimed review teaches recall, not selection
Reviewing answers slowly builds knowledge, which matters. It does not build the discrimination skill of choosing between two options that both look right. That only comes from repeated forced choices under a clock.
Format familiarity is a real skill
Papers written to authentic exam style train you to spot the pivot phrase, the negative qualifier, and the shift between diagnosis and management questions. Loose-format banks do not.
Practising the right way
The single highest-yield change most candidates can make is to shift from untimed question review to timed, exam-style mock papers at least twice a week in the final six weeks. If you are preparing for the membership exams from the UAE, working through a bank of MRCP Part 1 practice MCQs in full timed blocks, then reviewing your reasoning on every wrong answer, will build both the knowledge and the technique that the traps above target.
Track two numbers per mock: your raw score, and the number of questions where you changed a correct answer to a wrong one on review. The second number is a direct measure of how well your technique is holding up under pressure. When it falls to zero or one per paper, you are ready.
Frequently asked questions
What is the difference between best-of-five and single-best-answer MCQs?
In a traditional single-best-answer question, four options are unambiguously wrong and one is right. In a best-of-five question, several options can be defended in isolation, and the task is to pick the one that best fits the full clinical picture and the specific wording of the stem.
This is why keyword-matching fails on best-of-five papers. You have to weigh options against the exact question being asked, not against your general knowledge of the topic.
How much time should I spend per question in a postgraduate MCQ paper?
Most postgraduate medical exams give roughly 90 seconds per item on average. Some questions will take 30 seconds because the answer is obvious once you read the stem; others will need two minutes because the vignette is long.
The rule of thumb is: if you have read the stem twice and still cannot narrow it to two options, flag it and move on. Coming back with fresh eyes usually works better than grinding.
Should I always change my answer if I have doubts on review?
No. Research on test-taking behaviour consistently shows that first instincts are correct more often than candidates believe, and that changed answers are more often changed from right to wrong than the reverse.
Only change an answer if you have identified a specific piece of information in the stem that you missed the first time. “It just feels wrong” is not enough.
How many mock exams should I sit before the real exam?
As a minimum, sit two or three full-length, timed mocks under real conditions in the final month, and several shorter timed blocks each week before that. The goal is to make the pacing and the four-step reading method automatic.
Untimed question review is still valuable for building knowledge, but it should sit alongside timed practice, not replace it.
Does the setting described in the stem actually matter?
Yes. “Presents to a primary health centre”, “presents to the emergency department”, and “is seen in outpatient clinic” imply different available investigations and different appropriate next steps. A CT scan may be the ideal answer in a tertiary hospital and the wrong answer in a rural clinic where the correct step is to stabilise and transfer.
Read the setting as part of the clinical information, not as scene-setting.
Are negative qualifiers like ‘least likely’ common in these exams?
Common enough that missing one will cost you real marks over a full paper. Any time you see the words “least”, “not”, “except” or “unlikely” in the stem, physically mark it, either mentally or on scratch paper if the interface allows it.
Under time pressure the brain defaults to reading positive questions. The qualifier has to be forced into working memory.
How do I stop earlier questions influencing my answers on later ones?
Between each item, take one deliberate breath and reset. Do not carry the topic of the previous question into the next stem. Best-of-five papers deliberately cluster topics and then break them, precisely to catch candidates who pattern-match on recent context.
Practise this in mocks. Reset is a habit, and it only sticks if you drill it under exam-like conditions.
Soccer lover, foodie, ukulelist. Eames fan and screen printer. Let’s chat.