Short answer first: a good flashcard asks one clear question with one clear answer. To make them from your notes, you go through the material, find the key concepts, and make one card per concept rather than cramming a whole paragraph onto a single card. It works, but it takes time, which is why many people never make the cards. The alternative is to let a tool generate them automatically from your notes.
Why do flashcards work?
Flashcards are not just old-fashioned rote learning. They force you into active recall, pulling the answer out of memory rather than just reading it again. That is one of the best-documented ways to learn. In Karpicke and Roediger's 2008 experiment, students who kept retrieving vocabulary remembered about 80 percent of it a week later; students who kept restudying it remembered about a third. Combine it with spaced repetition, where you repeat the cards at growing intervals, and you can revisit material instead of relying on one successful attempt. The best interval and workload vary.
The flashcard study most students have never heard of
Nate Kornell tested this with actual flashcards in a 2009 paper. Students learned vocabulary either as one large stack, so each card came round again after all the others (spacing), or as four small stacks studied one after the other (cramming). The large stack won for about 90 percent of the participants. The twist is that most of them believed the small stacks had worked better, because the cards felt easier while they were studying. The feeling of fluency and the fact of learning pull in opposite directions, and the research says trust the fact. The 2013 review of ten study techniques reached the same verdict: practice testing and spacing are the two techniques with the strongest evidence.
The question is not whether flashcards work. It is that good cards take time to make, and bad cards are close to wasted effort.
What makes a good flashcard?
Before you make the cards, stick to these principles:
- One question, one answer. A card with five bullet points on the back is not a flashcard, it is a note. Split it up.
- Ask about understanding, not just wording. "What does X mean?" is better than reciting a sentence verbatim. You want to be able to use the concept, not just recognise it.
- Keep it short. The shorter the card, the faster you can review it, and the more repetitions you fit in.
- One idea at a time. If you are unsure what the card is actually asking, it is too big.
Piotr Wozniak, the creator of SuperMemo, calls this the minimum information principle: formulate each item as simply as possible, because simple items are recalled faster and scheduled more accurately. These are practical design recommendations, not a guarantee that every brief card produces understanding.
Three card types that work
- Question and answer. "What does a research question specify?" on the front, "What the investigation aims to find out" on the back. The default for definitions and mechanisms.
- Cloze deletion. A sentence from your notes with the key term blanked: "A bibliography lists the ___ used in a piece of work." Fast to make, good for facts.
- Reverse cards. Term to definition and definition to term as two separate cards. Useful for vocabulary and formulas, where you need both directions.
How to make flashcards from notes, step by step
- Read through your notes and mark the concepts, definitions, and connections worth remembering. Not everything needs a card.
- Make one card per concept. The front is the question, the back is the short answer.
- Rephrase instead of copying. Write the question in your own words. The act of rephrasing is part of the learning.
- Test yourself and fix the cards. If a card keeps going wrong, check whether the source is accurate, the concept is understood and the prompt is clear. Split it if it tests several separable ideas.
A worked example
Here is an invented assignment instruction used only to demonstrate card design:
The project proposal must state a research question, explain the chosen method and identify one limitation. Submit it as a PDF by Friday at noon.
A single card asking you to recite that paragraph mixes academic requirements with submission logistics. Instead, make focused prompts:
- What must the proposal say the project will investigate? / A research question.
- What choice must the proposal explain? / The method.
- What qualification must the proposal identify? / One limitation.
Put the deadline and file format in your calendar or submission checklist. They need action, not weeks of memorisation. For a real course, check every answer against the actual assignment brief. If you used the Cornell method in the lecture, the questions in your cue column are already the fronts of these cards.
How many cards, and how often
Start with a manageable set from one topic and notice how long it takes to answer and check the cards. A fifteen-minute slot can be an experiment, not a target that every module must fit. The Anki manual explains active recall and spaced review; exact intervals depend on the software and its settings. If the queue grows beyond your available time, pause new cards and review which prompts matter. Put those return visits into a study schedule, alongside the longer tasks needed for exam preparation.
The common mistakes
People most often make these three mistakes when building flashcards from notes:
- Too much on one card. You cannot recall a whole paragraph precisely, so you end up "almost knowing it", which does not count.
- Copying verbatim. If you just transcribe the note, you test recognition, not understanding. Turn it into a question.
- Never repeating. Writing cards may help you process material, but an unused deck gives you no later retrieval practice. Repetition over time is the whole point.
Two more are worth naming. Making cards for everything, so the deck becomes a second copy of the notes and the daily queue becomes an hour. And flipping the card too early, which turns a recall test back into rereading. Make a real attempt before revealing the answer; allow longer for a more complex prompt.
Let AI make the cards for you
The honest reason many people never make flashcards is that it is tedious and adds time on top of the reading itself. That is exactly the piece of work AI can take off your plate.
Notibo can generate draft flashcards from permitted recordings and uploaded documents. Check a sample against the source before starting the deck, including terminology, conditions and any missing context. Recording-generated cards and spaced repetition are Pro features included in the trial. Individual generated cards cannot currently be edited, so do not assume you can split or rewrite a poor prompt inside Notibo. Keep a separate question list or use a tool with manual card editing when that is essential.
How Notibo generates the cards from a lecture or a slide deck, and how the review schedule works, is on flashcards from lectures and slides.
If you want to understand why it works, read more about active recall and spaced repetition. Choose the workflow that lets you verify the answers and practise the actual skills your course requires.