Ask a student what their question bank percentage is and most can tell you immediately. Ask what they learned from yesterday’s block and the answer is usually vaguer. That asymmetry is the whole problem.
The explanation is the product
You are not buying questions. Questions are cheap and available everywhere. You are buying the explanation that follows each one, the paragraphs that tell you why the answer is what it is, why the distractors were tempting, and what distinguishes the presentation from the three conditions it resembles.
That is the teaching material. The question is the delivery mechanism that makes you receptive to it, because you cannot skim past an explanation for something you just got wrong.
Judge a bank on its explanations. Everything else is secondary.
That is the whole premise of QBankly, which puts its weight into the explanation and the review loop rather than into question volume. Whether that suits you is worth testing on a trial.
Percentage correct is the wrong metric
Your percentage measures what you knew when you started the block. It says nothing about what you learned during it, which is the thing you were there for.
Two students can both score sixty percent. One reviews carefully, understands every error, and will score higher next week. The other notes the number and moves on. The percentages are identical and the trajectories are not.
Track something else: how many of the errors you can now explain, and whether the topics you reviewed last week are the topics you are now getting right.
The questions you got right
Some of your correct answers were guesses, or were right for the wrong reason. Left unreviewed, those are indistinguishable from genuine knowledge, and they will reappear on the exam as confident errors.
The habit that fixes it costs almost nothing: flag anything you were not sure about as you go, regardless of whether you got it right, and review those with the same care as your errors. Students who do this consistently find their real gap list is noticeably longer than their wrong-answer list.
Timing and mode
Use tutor mode while you are still learning a topic, immediate feedback while the reasoning is fresh is a much better teacher than feedback an hour later.
Switch to timed blocks as the exam approaches. Pacing and stamina are separate skills from knowledge, and they are trained by practising under the conditions you will actually face.
Move from system-specific blocks to random ones as you consolidate. Random is harder for a reason that matters: on the real exam nobody tells you which system a vignette belongs to, and working that out is part of the task.
How much to do
Enough to finish one bank thoroughly, with full review, before the exam. That is a more useful target than a daily question count, and it is the constraint that should shape your schedule.
Finishing one bank properly beats starting two. It is not close.
Reviewing a block properly
The word “review” covers a wide range of behaviour, most of it ineffective. Reading the explanation for questions you got wrong is the minimum, not the method.
For each error, answer three things. Which feature of the vignette should have pointed you to the right answer? Why was the option you chose plausible, what would have had to be different for it to be correct? And is this a fact you did not have, or a fact you had and did not deploy?
That last distinction determines what you do next, and it is invisible if you only read the explanation and move on.
For questions you got right but were unsure of, do the same thing. Those are the ones that will otherwise reappear on the exam as errors you have no warning about.
A rough time budget
A block of forty questions takes about an hour. A review of that block, done as above, takes ninety minutes or so at first and settles somewhere shorter as the habit forms.
If your review is consistently taking twenty minutes, you are not doing the thing described here. You are checking answers. That is a much cheaper activity and it is worth considerably less.
If you change one thing
Flag every question you were unsure of, including the ones you got right, and review those with the same care as your errors. That single habit surfaces the gaps your percentage is actively hiding from you.
Frequently asked questions
Tutor mode or timed mode?
Tutor mode early, timed mode as the exam approaches. Immediate feedback is better for learning; timed blocks are better for building the pacing and stamina you will need on the day.
Random or by system?
By system while you are still learning a topic, random once you are consolidating. Random blocks are harder because identifying the system is itself part of the task, which is exactly what the real exam requires.