Active recall: why testing yourself beats rereading

Active recall, a form of active learning, means producing the answer from memory instead of rereading it. Asking yourself a question, reciting a line of reasoning with your eyes closed, filling a blank page, answering a flashcard or an MCQ: in every case, it is your memory doing the work, not your eyes.

The difference with rereading is fundamental. Rereading is recognizing: the text is in front of you and feels familiar. On exam day, however, you have to recall, that is, retrieve the information without help. Recognizing and recalling are two distinct operations, and only the second truly prepares you for the exam.

Rereading creates an illusion of mastery. On the third reading of a chapter, everything seems clear, because everything is familiar. This ease is misleading: it measures how fluently you read, not your ability to reproduce.

Highlighting suffers from the same flaw when done on its own. Highlighting a whole paragraph requires no effort of selection or memory. Highlighting remains useful as a marking tool, to flag what deserves a card or a question, provided it is followed by real recall work. See annotating your PDF lecture notes.

The testing effect refers to the fact that testing yourself on a piece of information locks it in better than studying it again for the same amount of time. It has been known for a long time, but the study by Roediger and Karpicke (2006) made it famous.

In that study, students read short texts, then either reread them or tried to recall their content in writing, without correction. Tested a few minutes later, those who had reread did slightly better. But tested two days and then a week later, it was the group that had tested itself that retained clearly more. The participants who had reread, for their part, believed they were better prepared.

This result sums up the trap: rereading seems more effective in the moment, and self-testing is more effective at the deadline that counts. An exam rarely takes place five minutes after studying.

In 2013, John Dunlosky and his colleagues published a literature review on ten common learning techniques, rating their utility according to the strength of the evidence and the variety of situations in which they work.

The two techniques rated most useful are precisely the ones that flashcards in spaced repetition combine: testing yourself, and doing it at increasing intervals. The two most widespread methods among students, rereading and highlighting, are among the least useful.

Low utility does not mean useless: rereading a chapter to understand it before making your cards is normal. The problem is making it your main method.

The blank page, also known as blurting, is a form of self-testing that requires no tools. It suits well-structured courses: a history chapter, a metabolic pathway, a legal regime.

• Read or review part of the course, then close it.
• On a blank sheet, write down everything you remember: outline, definitions, diagrams, examples, without looking.
• Compare with the course, and fill in, in another color, everything that was missing or wrong.
• Go back over the gaps: targeted rereading, then a new blank page a few days later.
• Turn recurring gaps into flashcards, so they join your spaced reviews.

This method also reveals how your knowledge is organized: if you recall the details but not the outline, it's the outline you need to work on.

Self-testing only works if your judgment is accurate. A few rules:

• Formulate your answer before looking, out loud or in writing. "I knew that" after the fact doesn't count.
• A half-right answer is a wrong answer, unless the question only asked for that half.
• Note hesitation: an answer found after a long effort is not mastered like an immediate one.
• Test yourself on questions you didn't see the day before: a test too close in time measures short-term memory.
• Accept bad scores. They show where to work, that is their point.

Testing yourself is more uncomfortable than rereading, precisely because it exposes the gaps. That discomfort is the sign the method is working.

In Mnemesia, flashcards are self-tests by nature, scheduled by FSRS. In addition, the self-test graded out of 20 lets you quiz yourself on a deck whenever you want, like a quiz in class: you get a score, and this run does not affect the scheduling of your cards. So you can test yourself the day before an oral quiz or an exam without upsetting your spaced reviews.

Each self-test is kept in a history, so you can see whether a subject is improving from one week to the next. Timed MCQs, written by you or received from a classmate, round out the practice, and good flashcards are made right next to the annotated handout.

Mnemesia is free, with no account and no ads, and your data stays on your device: download the app.