Vestibular system · practice questions vs Anki

The vestibular block is the one topic where a green Anki deck lies to you.

A card tests a rule you stated. The exam tests that rule run forward on a patient you have never seen. On most topics the gap is small. On the vestibular system it is the whole grade.

Direct answer · verified 2026-05-16

For the vestibular system, drill practice questions and keep Anki for only the flat-fact layer: the names of the canals and otolith organs, the planes they sense, the letters in COWS, labeled-diagram anatomy. Everything else in the block is chained directional reasoning, and a flashcard cannot test a chain. The honest split for this one topic is roughly ninety percent questions, ten percent cards. Caloric-test physiology here is checked against the standard account in NCBI StatPearls (Caloric Testing), re-verified 2026-05-16.

One fact, two treatments

Take caloric testing, the COWS topic. Here is that fact as an Anki card, and the same fact as a practice question. They are not testing the same skill, and only one of them matches what the exam does.

As an Anki card

Front

Caloric testing: what does COWS stand for?

Back

Cold Opposite, Warm Same.

What it tests: recall of four letters. By the time the card is mature you answer it in under a second. You were never asked to do anything with the rule.

As a practice question

During bithermal caloric testing, 30 degree Celsius water is irrigated into a patient's left ear. The vestibular system is intact. The fast phase of the resulting nystagmus beats toward which side?

A) Left
B) Right ← correct
C) Upward
D) No nystagmus is produced

What it tests: recall COWS, recognize 30 degrees as cold, apply Cold Opposite, then map opposite-of-left to right. Four linked steps. Break any one and you miss the item.

The chain the vignette makes you run

COWS is a shortcut over a six-step physiology chain. The flashcard hands you the shortcut. The practice question makes you walk the chain, which is the part that actually transfers to a question your professor worded differently.

Cold water, left ear: why the fast phase beats right

30 degree Celsius water enters the left ear canal

A cold caloric stimulus. Body core is 37 degrees, so 30 degree water cools the left labyrinth.

Endolymph in the left horizontal canal cools and grows denser

The denser column of fluid sinks under gravity.

The cupula deflects ampullofugally

Endolymph flows away from the ampulla, bending the cupula away from the utricle.

The left horizontal canal is inhibited

By Ewald's first law, ampullofugal flow in the horizontal canal lowers afferent firing on that side.

The brainstem reads a head turn toward the right

Reduced left-side input looks like the head rotating toward the intact right side.

Slow phase drifts left, fast phase beats right

The vestibulo-ocular reflex drives a slow drift toward the cold ear; the corrective saccade beats away from it. Cold Opposite.

A card that says “cold caloric causes ampullofugal cupular deflection” tests step three in isolation. It never asks you to carry step three through to step six, which is the only step the exam grades.

Why this block breaks the “just use both” advice

Most of medical school is recall with a thin application layer on top. Microbiology is mostly which bug, which toxin, which stain. Pharmacology is mostly which drug, which receptor, which side effect. For topics like that, a mature Anki deck does roughly eighty percent of the work, because eighty percent of what the exam asks is a fact you either retrieved or did not. Practice questions clean up the last twenty percent.

The vestibular system inverts that ratio. The flat facts are small and you can learn them in an afternoon: three semicircular canals, two otolith organs, the canals sense angular acceleration, the otoliths sense linear acceleration and head tilt. After that, almost every exam item is a chain. Which ear. Which canal. Which direction the endolymph moves. Excitatory or inhibitory. Slow phase or fast phase. Peripheral or central. The fact layer is maybe twenty percent of the block and the chained-reasoning layer is the other eighty.

That is why a vestibular Anki deck can be entirely mature and still leave you stuck on the exam. Maturity on a static card measures how fast you recognize the front, not whether you can run a six-step derivation under time pressure. The card "cold caloric causes ampullofugal cupular deflection" can be green in your deck for a month while you still cannot say which way the fast phase beats when the question names a specific ear and a specific water temperature. The card tested the rule. The exam tests the rule used.

So the honest split for this one block is not fifty-fifty. It is roughly ninety percent practice questions, ten percent Anki. Use cards for the handful of facts that genuinely have no chain attached, and spend the rest of your vestibular study time answering questions that make you derive a direction.

Anchor fact · auto-rephrasing on revisit

The stem rotates every revisit, so the answer slot moves and there is nothing to pattern-match

Here is the part no general "Anki vs questions" article covers, because it is a product behavior, not a study tip. studyly auto-rephrases the question stem every single time the item resurfaces. The underlying fact, the correct answer, the distractor pool, and the spaced-repetition schedule on that item all stay fixed. Only the surface wording rotates.

On a caloric-testing item that means revisit one shows you 30 degree Celsius water in the left ear, and the fast phase beats right. Revisit two, ten days later, shows you 44 degree Celsius water in the left ear, and now the fast phase beats left. Same anatomy, same lecture slide, same scheduled card. The answer slot moved, so there is nothing to pattern-match. You re-run Cold Opposite Warm Same from scratch, which is exactly what the exam will make you do.

Studyly · question-quality eval

Turbolearn · same documents

The questions are generated against the vestibular lecture your professor actually uploaded, not pulled from a generic web question bank written for a different curriculum, and not authored by ChatGPT with no quality rubric. On a held-out three-document eval scored blind on factual correctness, clarity, distractor quality, and question-type coverage, studyly-generated questions scored 81.3 of 100. Turbolearn scored 57.8 on the same documents; the rest of the field landed at 78.0 and 68.0. When a question item gets a vestibular direction wrong, the gap shows up on the exam as a wrong answer, so the rubric is the thing doing the work.

Side by side, on vestibular content specifically

This is not a general feature matrix. Every row is something the vestibular block specifically demands, where a static card and a rotating practice question behave differently.

An Anki card on a vestibular fact vs a studyly practice question generated from your lecture deck.

Feature	Anki card	Studyly practice questions
What a single item tests	Recall of one stated rule, for example that cold caloric stimulus causes ampullofugal cupular deflection.	The rule applied. Predict the fast-phase direction in a named patient, the way the exam phrases it.
The same item on revisit four	Identical front. You recognize the wording in under a second, so retrieval of the chain is no longer being measured.	Stem auto-rephrased: different ear, different water temperature. You re-run the chain on every pass.
Directional and spatial reasoning	Not exercised. A card states a direction, it never makes you derive one.	Case-style stems force the derivation: endolymph movement to cupula to canal firing to slow phase to fast phase.
A wrong answer	You press Again. No feedback on which link in the chain you actually broke.	Explain-my-mistake opens the exact slide from your lecture and names the step you skipped.
Building the material	Hand-typing cards from a 70-slide neuro deck, an hour or two, and you tend to skip the labeled-diagram slides.	Upload the deck, get roughly 200 questions across four formats in about 60 seconds.
Where the questions come from	Whatever you typed, or a community deck written for a different school's curriculum.	Your professor's actual vestibular lecture, the slides your exam is written from.

When Anki is still the right tool here

Anki is not the wrong tool for the vestibular system. It is the wrong tool for most of the vestibular system. There is a real flat-fact layer here and cards handle it well: the names of the otolith organs, which canal sits in which plane, the fact that the utricle senses horizontal linear acceleration and the saccule senses vertical, the literal letters in COWS, the structures on a labeled inner-ear diagram. None of those have a chain. They are pure retrieval, and pure retrieval is what spaced repetition was built for.

If your school has an upperclassman-maintained class deck for the neuro block, use it for that fact layer. If you already drill AnKing or Zanki for boards, leave it running. The argument on this page is narrow: do not let a green vestibular deck convince you the block is handled. The deck handles the facts. The directional questions are a separate skill, and on this block they are most of the grade.

studyly exports an Anki .apkg if you want one review queue, including image-occlusion cards over labeled diagrams, which are genuinely good for inner-ear anatomy. The one thing a static .apkg loses is the auto-rephrasing: once a question is a fixed Anki field, the stem stops rotating. So the practical pattern is to drill the case-style directional items inside studyly where the stem rotates, and export the flat-fact and image-occlusion cards to Anki where rote wording is the point.

Drill the chain, not the card front

Upload your vestibular lecture and get case-style questions that rephrase on every revisit. About 60 seconds, free tier on app.jungleai.com, no credit card.

Vestibular practice questions vs Anki: common questions

Practice questions or Anki for the vestibular system?

Practice questions, for almost all of it. The vestibular block is chained directional reasoning, and an Anki card tests a stated rule while the exam tests that rule applied to a specific patient. Keep Anki for the small flat-fact layer only: the names of the canals and otolith organs, which structure senses which plane, the literal letters in COWS, labeled-diagram anatomy. Spend the rest of your vestibular study time answering questions that make you derive a direction.

Isn't the standard advice to use both, since they're complementary?

Yes, and that advice is correct in general. Anki builds the foundation of facts, practice questions apply and refine them. The point of this page is not to reject that, it is to give you the ratio. On a recall-heavy topic like microbiology the split is roughly even because most of the exam is facts. On the vestibular system the flat-fact layer is small and the chained-reasoning layer is most of the block, so the honest split shifts to roughly ninety percent questions and ten percent cards.

Why do my mature vestibular Anki cards not translate to the exam?

Because a mature card means you recognize the front quickly, not that you can run a derivation. Maturity on a static card measures recognition latency. A vestibular exam item gives you a named ear, a named canal or a named water temperature, and asks you to chain endolymph movement to cupular deflection to canal firing to slow phase to fast phase. The card 'cold caloric causes ampullofugal deflection' can be green for a month while you still cannot complete that chain under time pressure.

What should I still make Anki cards for in the vestibular block?

The facts with no chain attached. The three semicircular canals and the planes they sense, the utricle versus saccule split for horizontal versus vertical linear acceleration, the literal expansion of COWS, the structures on a labeled inner-ear or cross-section diagram. Those are pure retrieval and spaced repetition handles them well. Image-occlusion cards over labeled anatomy diagrams are a particularly good fit and studyly can generate them for you.

Does generating vestibular questions with ChatGPT solve this?

Partly. ChatGPT will write you questions, but it enforces no quality rubric, does not track which items you got wrong, does not hold the answer key fixed while it rotates the wording on revisit, and does not know your professor's actual slides. On a held-out three-document eval scored blind on factual correctness, clarity, distractor quality, and question-type coverage, studyly-generated questions scored 81.3 of 100 against Turbolearn's 57.8. On a directional topic a weak distractor or an ambiguous stem is the difference between a question that trains the chain and one that teaches you the wrong direction.

What is auto-rephrasing and why does it matter for vestibular questions?

Auto-rephrasing rewrites the question stem every time the item resurfaces while the underlying fact and the correct answer stay fixed. On a caloric-testing item, revisit one might show 30 degree water in the left ear with the fast phase beating right, and revisit two might show 44 degree water in the left ear with the fast phase beating left. The answer slot moves, so you cannot memorize 'the answer is B'. You re-derive Cold Opposite Warm Same each time, which is what the exam forces. Static flashcards cannot do this because the front is a fixed field.

Does this logic apply to BPPV, Dix-Hallpike, and central versus peripheral nystagmus?

Yes. The whole vestibular block is directional. Dix-Hallpike asks which canal is involved and which way the torsional nystagmus beats. Localizing nystagmus asks you to chain features like direction-changing versus direction-fixed, suppression with fixation, and vertical versus horizontal into a peripheral or central call. Every one of those is a chain, not a flat fact, so the same conclusion holds: drill them as questions, not as cards.

Is there a free tier, and do I need a credit card?

There is a free tier on app.jungleai.com and no credit card is required to start. You can upload your vestibular lecture, generate questions across multiple-choice, free-response, case-style, and image-occlusion formats, and drill them without entering a card anywhere in the flow. Paid is opt-in.