Home › Editorial policy
How CoStudy decides what goes into a question bank, what gets corrected, and what we refuse to claim.
Every bank is built from the certifying body's published exam content outline — the document that lists the domains, their percentage weights, the question count and the passing standard. Chapters map to those published domains, and the number of questions in each chapter is proportional to the domain's published weight. Where a certifying body publishes no outline, the bank follows the most authoritative public syllabus available and says so.
We are not the certifying body, we have no access to live exam items, and we do not reproduce them. Any question bank claiming to contain real exam questions is describing something between a copyright violation and a lie.
Certification content is full of figures that are re-indexed every year — reporting thresholds, fee schedules, contribution limits, loan limits, continuing-education hour counts. A study bank that states last year's number as this year's rule teaches a candidate something false and does it confidently.
So our authoring rule is explicit: do not state a threshold, dollar amount, hour count or deadline we are not confident is current; where a question would require one, write a different question that tests a principle we are certain of. Items test which rule applies, what the sequence is, who bears the duty and what the consequence is. Where a scenario needs a variable figure to work, the figure is supplied inside the scenario as a given rather than asserted as a rule.
Two mechanical biases let someone score well on a practice bank without knowing the material, and both are easy to introduce by accident when writing at volume. The first is answer-letter clustering: if a disproportionate share of correct answers are B, a test-taker learns to guess B. We measure the letter distribution across every bank and rebalance until no letter is meaningfully over-represented.
The second is length bias. Writers naturally give the correct answer more detail than the distractors, so "pick the longest option" becomes a winning strategy that teaches nothing. We measure how often the correct answer is the longest and the shortest choice, and rewrite until neither is far from chance. We also verify that every choice is correctly lettered and ordered, that no question is duplicated or near-duplicated within a bank, and that every multiple-choice item carries a rationale addressing all four options.
Exams get renumbered, retired and reweighted, and a bank written to a superseded outline teaches the wrong proportions even when every individual question is correct. Before writing or revising a bank we re-check the current outline rather than assuming the version we last saw is still live. When a weighting changes we redistribute the bank and record what the previous weighting was, so the change is visible rather than silently overwritten.
If you find a question that is wrong, out of date or ambiguous, tell us and we will fix it. A catalogue of this size will contain errors; the useful question is how quickly they are corrected once found. Material corrections are recorded in the corrections log.
Bank pages carry a byline naming who maintains them. Where a bank has been reviewed by a credentialed subject-matter expert, the byline says so and names the reviewer with their credentials; where it has not, it says "written and maintained by" instead. We do not describe maintenance as review. In regulated fields — nursing, accounting, securities — treat our material as practice and check anything consequential against the certifying body's own published guidance.
How our questions are written → · Corrections log · Browse the banks