SampQs is a CCFP SAMPs bank whose "AI tutor" is, in practice, an automated marker: you write a free-text answer and it scores it against a guideline-based marking scheme, showing the marks you earned and missed. That is genuinely useful, but it is not a conversational tutor you can ask "why," so the reasoning has to stay yours. This workflow is for family medicine candidates using SampQs for the written SAMP component who want the AI to sharpen their thinking, not replace it. Its limitation: it does not rehearse the oral.
What SampQs offers for CCFP right now
Figures below are vendor-reported from sampqs.com and were last checked on 19 July 2026. Confirm current numbers and pricing on the product page before you subscribe.
| Item | Detail (vendor-reported, last checked 19 July 2026) |
|---|---|
| Exam covered | CCFP written — SAMPs only (no SOO, no MCQ component) |
| Volume | 2,400+ SAMP questions across 598+ clinical cases; 105 priority topics |
| "AI tutor" | Automated grading of free-text answers against a marking scheme; shows marks earned and missed and the expected answers with guideline references |
| Adaptive | Tracks performance and resurfaces weak topics; readiness dashboard |
| Mock exams | 12-case, two-hour simulations |
| Access / price | 30 free practice cases without a card; subscription pricing on the pricing page (verify current tiers) |
| Not reproduced | The live SOO; the MCQ and short-menu format arriving in the written exam from 2026 |
The important nuance, and the reason for this article's title: SampQs' AI grades, it does not converse. It will tell you that your answer missed "screen for retinopathy," but it will not, on its own, walk you through why or let you counter-argue. Where this workflow calls for open-ended interrogation, you supply it — as self-questioning against the marking scheme, or by taking the item to a genuinely promptable tutor.
The CCFP exam you are actually preparing for
The CFPC Certification Examination in Family Medicine has a computer-based SAMP written component of four hours and an oral component of five 15-minute Simulated Office Orals on a separate day. The College does not publish a fixed per-component question count, so verify current figures on cfpc.ca.
The format change matters here more than for most banks. From April 2026 the SAMP moves from written free-text toward multiple-choice and short-menu items — up to 25 per cent of Fall 2026 cases, all cases from 2027 (confirm the proportion for your diet on cfpc.ca). SampQs' free-text grading trains exactly the reasoning the exam rewards, but the answer format it drills is the one being phased down. Use it for reasoning and content; get format rehearsal from official and MCQ-format material.
A reusable interrogation taxonomy
Because SampQs marks rather than converses, you need a fixed set of questions to interrogate each miss yourself. Six categories cover almost everything:
- Mechanism: why does this intervention work, physiologically or pharmacologically?
- Discriminating features: what one or two findings separate this diagnosis from its nearest mimic?
- Option elimination: why is each distractor wrong, not merely why the key is right?
- Guideline verification: which Canadian guideline says this, and what edition or year?
- Counterfactual: change one variable in the stem — what changes in the answer?
- Retrieval test: can I reproduce the marking points cold, without the scheme in front of me?
Twenty-five prompts, grouped by the six common error types
Keep these reusable and free of any copied question text. Pose them to yourself against the marking scheme, or to a promptable tutor where you need a genuine back-and-forth.
Knowledge gap: (1) What is the single Canadian guideline that governs this? (2) State the threshold or target number exactly. (3) What is the first-line agent and the first alternative? (4) Which finding here is the diagnostic anchor?
Misread stem: (5) Which qualifier in the stem did I skip? (6) Re-read for age, pregnancy, renal function and allergies — did any change the answer? (7) What is the actual task verb — diagnose, investigate or manage? (8) Is this the first step or a later one?
Premature closure: (9) What are the next two diagnoses I did not consider? (10) What would have to be true for my first answer to be wrong? (11) What red flag would change urgency? (12) Did I stop generating too early?
Guideline localisation: (13) Is this recommendation Canadian, and from which body? (14) What year is the guidance, and has it been superseded? (15) Where does Canadian practice differ from US or UK guidance here? (16) Does provincial variation affect the answer?
Management sequencing: (17) What must happen before the step I chose? (18) What is the safest first action if I am uncertain? (19) What is the disposition — treat, refer, admit or safety-net? (20) What follow-up interval does the guideline set?
Retention: (21) Can I list every marking point from memory now? (22) Can I still do so in a week? (23) Can I answer a fresh case on the same principle? (24) What one sentence captures the rule? (25) When is this item next due for review?
The anti-answer-leak rule
Commit before you reveal. Write your full free-text answer and a one-line rationale, submit it, and only then read the marking scheme. If you glance at the expected answer first, you convert an active-recall task into passive reading and lose most of the benefit. With a grader this is easy to enforce, because the scheme is hidden until you submit — do not defeat that by pre-reading a model answer elsewhere.
A jurisdiction check you run every time
Because guideline localisation is where family-medicine candidates lose defensible marks, keep one prompt permanent: "Does this marking point reflect current Canadian guidance, from which body, and of what year?" Run it against the scheme's cited references, and where the scheme is silent, verify against the primary Canadian source yourself. This single habit catches the most common and most costly SAMP error.
Keep a misconception record
For each substantive miss, write four lines: the incorrect rule you held; the corrected rule; one transfer question that tests the corrected rule; and a review date. Four lines, not a transcript. The transfer question is the point — it forces you to prove the correction on a fresh item rather than re-reading the marking scheme.
A weekly verification sample
Automated grading against a marking scheme is only as current as the scheme. Once a week, take a small random set of SampQs items — five is plenty — and independently check the marking points against the primary Canadian source. Log any discrepancy: an outdated threshold, a superseded first-line agent, a guideline that has moved. This protects you from silently learning a stale rule, and it is the same discipline the framework pillar on calibrating AI-graded feedback recommends before you trust any automated score.
Worked example: a seven-day week around clinical work
Give SampQs one job — AI-graded SAMP reasoning — and iatroX one job — transfer practice on fresh, unseen items. No proprietary-algorithm claims are needed; a second non-overlapping bank simply supplies clean stems.
- Monday: 6–8 SampQs cases, free-text, committed before reveal. Interrogate each miss with the six-category taxonomy.
- Tuesday: for the principles you missed, answer fresh iatroX items on the same principles; right cold means retire, wrong means step back.
- Wednesday: 6–8 new SampQs cases. Run the jurisdiction check on every guideline call.
- Thursday: your weekly five-item verification sample against primary Canadian sources; update the misconception record.
- Friday: spaced retrieval of last week's corrected rules, from memory.
- Saturday: a 12-case, timed SampQs mock; read pacing and unseen first-attempt accuracy.
- Sunday: rest or a short error-log review; pick next week's weakest topic.
Decision checklist: continue, supplement, switch or stop
- Continue while SampQs misses teach new content and your verified accuracy is rising.
- Supplement with iatroX for unseen transfer once misses become "I have seen this scheme" rather than "I did not know this."
- Switch only where a blueprint audit shows a domain SampQs does not serve and another resource does.
- Stop new questions when coverage, unseen timed performance, pacing, retention and official-material calibration are all met — then move to official material and SOO rehearsal.
Bottom line
SampQs' AI-graded SAMP practice is a strong reasoning-and-content resource, and free-text grading against a guideline scheme is a genuinely good way to expose sloppy thinking. Just be clear about what it is: a marker, not a tutor. Keep the reasoning yours, commit before you reveal, verify the scheme against primary Canadian sources weekly, and prove every correction on a fresh item. Do that and the AI sharpens your clinical reasoning instead of quietly doing it for you.
Frequently asked questions
Is SampQs enough for CCFP on its own? For the written SAMP reasoning it is a substantial resource, and its free-text grading trains a skill many banks cannot. But it covers SAMPs only — no SOO and, at present, no MCQ-format practice — so it cannot be your whole stack. Pair it with official CFPC material for calibration, spoken practice for the oral, and a second bank for unseen transfer, and it becomes a strong component rather than a complete preparation.
Which CCFP component does SampQs not reproduce well? The Simulated Office Oral, which it does not attempt at all. SampQs is a written-SAMP grader; the SOO is a live, verbal, timed performance with a standardised patient, assessing communication and reasoning in real time. No free-text bank rehearses that. SampQs can build the underlying knowledge an SOO draws on, but the oral itself needs spoken rehearsal with a partner or a recognised course.
How should I verify SampQs AI answers for CCFP? Treat the marking scheme as a claim to be checked, not gospel. Each week, take a small random sample of items and confirm the marking points against the primary Canadian source — Hypertension Canada, Diabetes Canada, the CTFPHC, the SOGC or the CFPC's own guidance — and log any threshold, first-line agent or recommendation that has moved. Run a standing jurisdiction prompt on every guideline call. This catches outdated or non-Canadian marking points before you memorise them.
When should I stop using SampQs and move to mixed mocks? Move to full-length, timed mocks per topic as each clears its first-attempt accuracy floor on unseen items, rather than waiting to finish the bank. Stop adding new SampQs cases when coverage is complete, unseen timed performance is stable, pacing is comfortable and your weekly verification is turning up nothing new — at which point official material and oral rehearsal give a better return than more free-text cases.
How should I combine SampQs with iatroX without duplicating practice? Assign non-overlapping jobs. SampQs is your AI-graded SAMP-reasoning bank; iatroX is your unseen-measurement and transfer layer. When SampQs marks you down on a principle, do not re-do that case — commit to a rationale on a fresh iatroX item testing the same principle, using the Socratic prompt before you reveal the explanation. That keeps your measurement uncontaminated and turns two resources into a verify-then-transfer loop. The two-Q-bank rule sets out how to avoid overlap.
Editorial notes and references
Written by Dr Kolawole Tytler, NHS GP and founder of iatroX. Last checked 19 July 2026; all platform figures are vendor-reported and may change — verify the current count, features and pricing on the product page. This article characterises SampQs' AI honestly as an automated free-text grader rather than a conversational tutor, based on the vendor's own description as of the check date. Disclosure: iatroX operates a CCFP question bank and a Socratic Tutor that compete with SampQs; iatroX's role here is confined to unseen transfer practice and verification, jobs SampQs' grader does not claim, and it is not presented as an SOO simulator. Corrections via the feedback route on iatrox.com.
References: CFPC — Certification Examination in Family Medicine (cfpc.ca); SampQs product and pricing pages (sampqs.com); iatroX CCFP bank (https://www.iatrox.com/canada/exam/ca-ccfp); "Calibrating AI-graded SAQs and OSCEs" (https://www.iatrox.com/blog/ai-graded-saqs-and-osces-how-to-calibrate-automated-feedback-before-you-trust-the-score); "Your Q-Bank Percentage Is Not Your Exam Score" (https://www.iatrox.com/blog/qbank-percentage-not-your-exam-score).