How to Turn Notes Into Multiple Choice Questions
To turn notes into multiple choice questions, you take one testable claim, make that claim the correct answer, and build the wrong options out of mistakes a person could plausibly make about it. Almost all of the work sits in the wrong options. A question with one right answer and three obviously silly ones is not a test, it is a reading comprehension exercise with extra steps.
That is the whole concept. The rest of this piece is about what "plausible" means, where those wrong options come from, and what multiple choice cannot do for you no matter how well you write it.
A multiple choice question is a retrieval prompt with competition
When you read a note, your brain recognizes it. Recognition feels like knowing. Retrieval is different: you have to pull the answer out with nothing in front of you, and that act of pulling is what strengthens the memory. This is the testing effect, documented across decades of work by researchers including Henry Roediger and Jeffrey Karpicke.
Multiple choice sits in an awkward middle. The correct answer is right there on the screen, so the format leans toward recognition. But it does not have to stay there. In a study by Little, Bjork, Bjork and Angello published in Psychological Science, multiple choice questions with genuinely competitive alternatives produced better later recall than straightforward cued recall questions, including for information that was never directly tested. The mechanism they proposed is simple: when every option looks possible, you cannot just recognize the right one. You have to retrieve enough about each wrong one to rule it out. Four small retrievals instead of one.
Weak distractors kill that effect completely. So the design rule follows directly: a wrong answer earns its place only if you have to actively reject it.
Step one: find the claims, not the topics
Open your notes and look for sentences that can be true or false. Those are your raw material. Headings, definitions written as fragments, and diagrams without labels are not testable yet.
A useful filter: if you cannot imagine someone getting it wrong in a specific way, skip it for now. "Photosynthesis" is a topic. "Photosynthesis produces oxygen as a byproduct of splitting water, not of fixing carbon dioxide" is a claim, and it comes with its own built in wrong answer.
From a typical page of lecture notes you will find maybe five or six real claims. That is a normal yield. Trying to force twenty questions out of one page is how you end up with trivia about which slide number something appeared on.
Step two: write a stem that stands alone
The stem is the question part. The test of a good stem: cover the options with your hand and try to answer it. If you cannot, the stem is incomplete and the options are doing work they should not be doing.
Three habits that quietly ruin stems:
- Grammar leaks. "The organelle responsible for ATP synthesis is an..." tells the reader the answer starts with a vowel. Rewrite as "Which organelle is responsible for ATP synthesis?"
- Length leaks. Writers tend to over qualify the correct answer to make it defensible, so the longest option is usually right. Trim the key or pad the distractors until they match.
- Negative stems. "Which of the following is NOT..." tests careful reading more than knowledge. Use them rarely, and capitalize the negation when you do.
Step three: build the distractors from real confusions
This is the part that separates a useful question from filler. Your wrong answers should come from somewhere specific, and your own notes usually contain the source.
| Where the distractor comes from | What to look for in your notes | Example |
|---|---|---|
| The neighboring concept | Two terms introduced in the same paragraph | Mitosis vs meiosis, weight vs mass |
| The reversed relationship | Any sentence with a direction or causal arrow | "Insulin raises blood glucose" for a question whose answer is "lowers" |
| The right idea, wrong number | Ratios, dates, counts, thresholds | 2 sodium out and 3 potassium in, when the answer is 3 out and 2 in |
| The common misconception | Anything your teacher explicitly warned about | "Deoxygenated blood is blue" |
| The correct answer to a different question | Adjacent facts from the same lecture | Naming the enzyme when the question asks for the substrate |
Notice that every row points back at material you already have. You are not inventing wrong answers, you are collecting the ways the material is genuinely confusable.
A worked example. Take this note: "The sodium potassium pump moves 3 Na⁺ out of the cell and 2 K⁺ in, using one ATP per cycle."
How many ions does the sodium potassium pump move per ATP hydrolyzed? A. 3 Na⁺ out, 2 K⁺ in B. 2 Na⁺ out, 3 K⁺ in C. 3 Na⁺ in, 2 K⁺ out D. 3 Na⁺ out, 3 K⁺ in
B swaps the numbers, C reverses the direction, D removes the charge imbalance that makes the pump electrogenic. Each wrong option corresponds to a real thing students mix up, and rejecting each one requires you to know something. That is a question worth answering.
Doing this at volume
Writing questions this carefully takes time, which is the honest catch. Ten good questions from a chapter is maybe twenty minutes of work, and that work is itself excellent studying. Many students find that writing the questions teaches them more than answering them later.
The tradeoff is real though, and it is why automation exists. Qora is an iOS app built around this specific step: you give it your own material, a photo of the page, a recording of the lecture, or text you typed, and it produces a short sectioned lesson plus questions drawn from that material, including multiple choice, fill in the blank, written answer and open recall. Because the questions come from your notes rather than a shared question bank, the distractors are drawn from the same content you are studying. The text recognition and transcription run on the device, so photos and audio stay on your phone, and there is no account to create.
Whether you write them yourself or generate them, the quality check is the same: read the four options and ask whether the wrong ones cost you anything to reject.
Where multiple choice stops helping
Three limits worth taking seriously.
Recognition is not production. If your exam asks you to derive, explain or write, practicing only multiple choice will overstate how ready you are. The fix is not to abandon the format, it is to mix it with free recall, where you close the notes and write everything you can remember about a topic on a blank page.
Wrong answers can stick. Roediger and Marsh showed that multiple choice testing has a negative side: readers sometimes later reproduce the lures they saw as if they were true. The protection is immediate feedback. Never answer a set of questions without seeing the correct answer right after, and read the explanation for the ones you got wrong rather than just noting the score.
Some material resists the format. Proofs, essay arguments, long calculations and anything requiring you to weigh evidence do not compress into four options without losing the thing being tested. Do not force it. Use written answer questions there, or just work the problem.
Start with one page of notes tonight. Find five sentences that could be true or false, turn each into a question with three distractors pulled from the table above, and put the page away before you answer them tomorrow. The questions you struggle to write distractors for are usually the parts you understand least well, which makes the writing itself a diagnostic. If the manual pass convinces you the format is worth using but the volume is the obstacle, that is the point at which a tool like Qora earns its place.