This workflow is for the core psychiatry trainee who wants to use MRCPsychMentor's strength-and-weakness signals for MRCPsych Paper B without letting a rising overall score hide the critical-review and statistics third that self-selected practice reliably under-serves. It addresses the written MCQ-and-EMI paper, not the CASC. The principal limitation to design around: the platform scores the topics you choose, so "adaptive" study is only as honest as the floors you set — and for Paper B the floor that matters most is the statistics one.
What MRCPsychMentor offers for MRCPsych Paper B right now
Vendor-reported, observed 20 July 2026 — verify on the product page before relying on any figure. MRCPsychMentor lists over 2,000 Paper B questions (best-answer one-of-five MCQs and EMIs) curated from previous exams and the current College syllabus across sections 6–14, including research methods, statistics, critical review and evidence-based practice. It offers a daily-updated histogram comparing your score against other candidates, strength-and-weakness analysis, question-review functions, revision notes under each question, two mocks structured to the College's category distribution, and revision and timed modes. Access is 4 months for £60 and 6 months for £80, with a free demo. On the pages we checked it does not publish a proprietary adaptive-difficulty algorithm, a national percentile or a predicted pass score, so this workflow builds the "adaptive" behaviour deliberately and makes no proprietary-algorithm claims. A separate AI-powered CASC resource exists but is not part of the Paper B bank.
The exam you are training for
MRCPsych Paper B is 150 questions in three hours at Pearson VUE — about two-thirds best-of-five MCQ and one-third EMI, at a pace near 72 seconds per item. Crucially, one third of the paper — roughly 50 marks — is critical review (research methods, statistics, appraisal and evidence-based practice), and two thirds is clinical, across general adult, old age, child and adolescent, forensic, addictions, psychotherapy, learning disability and service organisation. No negative marking. The clinical CASC is a separate exam. Because the critical-review weighting is fixed and large, the whole workflow below is built to stop that third being crowded out by the clinical topics candidates find easier.
Baseline week: measure before you personalise
Before you let any strength-and-weakness feature shape your study, complete a small, blueprint-stratified unseen sample. Draw a proportional spread — importantly, a full third of it from critical review and statistics — timed, mixed and sat without notes, around 90 items in total. Record first-attempt accuracy separately for the clinical and critical-review halves. This baseline is your reference point; a personalised feed that lacks it will only confirm you improved on what you practised. Take the baseline on a bank you will not learn from so it stays clean — an iatroX Paper B block is built for this measurement job.
First pass: set domain floors
Set a floor for every area before you begin, and make the critical-review floor non-negotiable. A floor is a minimum number of first-attempt items you will complete in a domain before allowing yourself to grind a favourite topic. Commit, for instance, to at least 300 first-attempt critical-review items — study design, bias and confounding, diagnostic-test statistics, number needed to treat, meta-analysis — plus floors for low-volume clinical areas such as learning disability and psychotherapy. Floors defeat the commonest failure of "adaptive" study: an overall score that rises precisely because you have stopped attempting the statistics third the exam guarantees will appear.
Error taxonomy: label every miss
Tag every wrong answer with one of six labels, because each implies a different fix:
- Knowledge gap — you did not know the concept. Fix with a short source read and a spaced re-test.
- Misread stem — you knew it but misread the question, data table or EMI options. Fix with technique.
- Premature closure — you chose before reading all options. Fix with a deliberate pause.
- Guideline error — you used outdated clinical guidance. Fix against the current source, taking medicines detail from the SmPC or electronic medicines compendium.
- Calculation error — you fumbled the statistics arithmetic (risk, likelihood ratios, confidence intervals). Fix by drilling that calculation type.
- Time-pressure error — you would have got it with more time, common on data-table items. Fix with timed practice.
Tagged errors over a fortnight tell you whether your Paper B problem is clinical knowledge, statistical reasoning, technique or pace — each with its own remedy.
Review interval: not everything deserves an immediate repeat
Decide the follow-up from the tag. A knowledge or statistics-concept gap deserves a short source read now and a fresh, different transfer item in a few days, not an immediate repeat that only trains recognition. A calculation error deserves a small block of similar calculations. A misread or premature-closure error deserves a technique note and no new content. Reserve immediate repeats for almost nothing, spaced re-tests for genuine knowledge gaps, and always prefer a new item on the same concept to re-answering a remembered one.
Mixed-block switch: when to stop filtering
Topic-filtered practice builds; timed mixed blocks test. Switch the balance towards mixed blocks when your floors — especially the critical-review floor — are met, your first-attempt accuracy in weak areas has stabilised, and the exam is a few weeks away. Make timed, random, blueprint-mixed blocks your main instrument, ensuring each block carries its proper third of statistics, and keep filtered practice only for any area still below floor.
Exit criteria: what "ready" actually means
You are ready to taper when you can evidence five things, none of which is bank completion: a coverage floor met across clinical and critical-review content; stable first-attempt performance on unseen mixed blocks, including the statistics sub-score; pacing inside roughly 72 seconds per item; retention on spaced re-testing; and official-material calibration against the College's sample questions and standard. "I finished the bank" is not on that list. The signals you can trust most come from unseen items the system did not choose for you.
Worked example: a seven-day plan around clinical work
For a trainee on a full rota, using MRCPsychMentor for building and iatroX for unseen measurement:
- Day 1 (Mon) — 40 critical-review and statistics items (the priority floor), tagged by error type.
- Day 2 (Tue) — 30 general-adult and old-age items; short source reads on two tagged knowledge gaps.
- Day 3 (Wed) — 30 child-and-adolescent and forensic items, then a 30-item unseen, timed, mixed iatroX block; log first-attempt accuracy, watching the critical-review sub-score.
- Day 4 (Thu) — 40 more critical-review items on the weakest sub-topics; re-test Monday's misses as fresh questions.
- Day 5 (Fri) — one timed MRCPsychMentor mock; review by error tag, not just score.
- Day 6 (Sat) — consolidation: spaced re-tests of the fortnight's statistics and knowledge gaps; two short readings.
- Day 7 (Sun) — a 60–90-item unseen, timed, mixed iatroX block; compare with Wednesday and the baseline to confirm the statistics third is transferring.
This is the two-bank discipline from The two-Q-bank rule: one bank to learn on, a separate unseen bank to measure on. For a deeper read of the analytics, see the companion MRCPsychMentor Paper B analytics audit.
Decision checklist: continue, supplement, switch or stop
| Signal (measurable, not sunk cost) | Action |
|---|---|
| Floors not yet met; critical-review accuracy still rising | Continue topic-filtered practice against the floors |
| Overall score high but unseen mixed first-attempt lags; critical review under floor | Supplement — force the statistics third, add unseen mixed blocks |
| Floors met, first-attempt stable, but pace on data items untested | Switch to timed mixed mocks |
| Unseen blocks confirm the standard; retention holding; errors now only technique | Stop drilling; taper and rest |
Frequently asked questions
Is MRCPsychMentor enough for MRCPsych Paper B on its own? For the written paper, over 2,000 MCQ and EMI items (vendor-reported, 20 July 2026) with category-weighted mocks can serve as a main bank for many candidates — but only if you set a hard critical-review floor and measure transfer somewhere unseen, because avoiding the statistics third is the classic way to pass the dashboard and fail the paper. It does nothing for the CASC, so pair it with the College's materials and a second bank.
Which MRCPsych Paper B component does MRCPsychMentor not reproduce well? The clinical CASC, which no written bank rehearses — the vendor's separate AI CASC tool needs its automated marking calibrated against a human before you trust it. Within the written paper, recognition items under-rehearse the end-to-end appraisal of an unfamiliar study, which is closer to what the critical-review third and clinical practice actually demand than picking a definition from five options.
How many MRCPsychMentor questions should I complete per day for MRCPsych Paper B? Spread matters more than volume: roughly 30–50 tagged, reviewed first-attempt items on weekdays with a standing rule that at least a third are critical review and statistics, more at weekends with a timed mock. Protect one unseen, timed, mixed block each week as your real signal, and let your floors, not a daily total, decide which topics the items come from.
When should I stop using MRCPsychMentor and move to mixed mocks? When your floors — critical review included — are met, your weak-area first-attempt accuracy has stabilised, and the exam is a few weeks away. Filtered drilling then yields diminishing returns and repeat accuracy flatters you; make timed, random, blueprint-mixed blocks that carry their proper statistics share your main instrument.
How should I combine MRCPsychMentor with iatroX without duplicating practice? Give each a single job: MRCPsychMentor to build and repair (first-pass learning, explanations, spaced re-tests, including the statistics items) and iatroX only to measure (fresh, unseen, timed, mixed blocks once or twice a week). Different item pools mean you are not re-answering the same questions; you are separating "have I learned it" from "can I retrieve it cold", which is the whole point of the two-bank rule.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 20 July 2026; MRCPsychMentor question counts, prices and features are vendor-reported as of that date and change without notice — verify the current figures on the product page. Disclosure: iatroX operates a competing MRCPsych Paper B question bank; here its role is confined to the unseen-measurement job MRCPsychMentor does not claim to perform, and this article takes no position on price. Corrections are welcome via the feedback route on iatrox.com. References: Royal College of Psychiatrists — MRCPsych examinations, syllabus and the critical-review weighting (rcpsych.ac.uk); MRCPsychMentor Paper B product page (mrcpsychmentor.com; vendor-reported); iatroX MRCPsych Paper B bank (https://www.iatrox.com/mrcpsych-paper-b); "Your Q-Bank Percentage Is Not Your Exam Score" (https://www.iatrox.com/blog/qbank-percentage-not-your-exam-score); "The two-Q-bank rule" (https://www.iatrox.com/blog/the-two-q-bank-rule-how-to-add-a-second-bank-without-duplicating-questions-or-destroying-calibration).