Active recall: how to study a recorded lecture

Active recall is studying by answering a question from memory before you look at the answer. You close the notes, say or write what you remember, then check. Repeated and spaced out, this one habit does most of the work that separates a lecture you can use from a lecture you merely attended.

Active recall — also called retrieval practice — is the act of pulling information out of your memory instead of putting it back in front of your eyes. The direction of the movement is the whole difference.

What active recall is

Rereading, highlighting and copying notes all place information in front of you and ask you to recognise it. None of them asks you to produce it. Recognition feels fast and comfortable. Production is slower, often uncomfortable, and that discomfort is the signal that memory is being worked rather than admired.

The students who ask how to apply this usually know the term already. What they are missing is a question with an answer short enough to check. Without that, "do active recall" is advice with no first step.

What the research says about active recall

Roediger and Karpicke (2006) gave students short prose passages to study, then had them either restudy the passages or take free-recall tests on them. Five minutes later, the students who had restudied recalled more. A week later, the students who had been tested recalled more. The restudy group also left more confident that they would remember.

That is the result to keep in mind. The method that wins on an immediate check-up is not the method that wins on the exam, and the feeling of being prepared ran opposite to being prepared.

The study did not change how much time students spent with the material. It changed what the time was spent on.

Why rereading feels like studying

A page you have read three times is easy to read a fourth time. That ease is fluency, and the brain reads fluency as familiarity, and familiarity as knowledge. Highlighting speeds the illusion up: the page comes back marked, so the session looks like output.

None of that is memory. It is a page that has been visited. If you want evidence, close the material now and write the three main points of the last lecture you attended. The gap between what you expected and what came out is the size of the problem.

Rereading is not worthless. In the same experiment it won when the test came five minutes later. If you need the material for the next hour, restudy it. If you need it in a week, test yourself.

How to turn a recorded lecture into questions

A recording is the wrong shape for recall, because nobody can answer three hours of audio. The transcript is the raw material: it is text, it can be searched, and it keeps the exact sentence you will check your answer against.

• Search for definitions. Phrases like "is", "we call this" and "defined as" mark them.
• Search for numbers, names and dates. These are what exams ask and what memory drops first.
• Turn each one into a single question with a one-sentence answer.
• Keep the timestamp beside the question, so checking takes seconds instead of minutes.
• Skip the stories, the housekeeping and the asides.

A lecture yields fewer questions than you expect, and that is good news. Twenty questions answered from memory beat two hours of rereading the same material.

A 20-minute active recall session

• Open the question list. Keep the transcript closed.
• Answer the first question out loud or in writing, in a complete sentence.
• Check against the transcript, not against whether the answer "sounds right".
• Mark it right, partly right or wrong.
• Continue to the end. Do not stop to relearn an answer you missed.
• After the last question, go back and re-answer every one you did not get clean.
• Finish on time. The next pass is in a few days, not tonight.

The first session is the slowest and the worst-feeling. That is expected. The second pass takes about half as long, the third less again.

How to space reviews before an exam

Two reviews in the same evening are one review with extra steps. The evidence on spacing is clear about direction and less exact about the number, so work with a range.

Cepeda and colleagues (2006) pooled 839 assessments from 317 experiments reported in 184 articles. Their conclusion: the interval that produces the best final retention grows as the delay before the test grows. A shorter gap is not automatically better. The right gap is the one matched to the date you need the material.

Cepeda and colleagues (2008) then taught more than 1,350 people, reviewed the material after a gap of up to 3.5 months, and tested them up to a year later. The best gap was about 20 to 40 percent of a one-week delay, and about 5 to 10 percent of a one-year delay.

What that means in practice:

• Test in 7 days: leave roughly 1.5 to 3 days between reviews.
• Test in 1 year: leave roughly 2.5 to 5 weeks between reviews.
• The longer the delay, the longer the gap — but the smaller its share of the delay.

Notice what the numbers rule out. Cramming the night before and again the next morning is what most students do for an exam a week away, and it is the opposite of the pattern the data points to.

How rereading, rewriting and recall compare

The third row is not better because it is harder. It is harder because it is better, and the difficulty is the part you can feel.

What to do with the answers you get wrong

Wrong answers are the output of the session, not its failure. They tell you which sentences have not stuck, and no amount of rereading will tell you that.

Mark them and re-answer them inside the same session. At the next spaced pass, start with them. Write the correct sentence from the transcript, close it, then reproduce the sentence from memory without looking.

Do not reread the whole section because of one wrong answer. That turns a five-minute correction into a forty-minute reread, and the reread is the habit this method exists to replace.

How to use active recall two days before the exam

Do not read the whole transcript again. Answer the questions you marked wrong, then go once through every question in the order the syllabus gives them. If an answer still fails, open the timestamp and hear the lecturer say it.

The days before an exam are too late to build understanding, but they are a good time to audit what is stored. A final pass that ends with the material still hard to recall is a working pass. A final pass that feels easy is usually a reread in disguise.

How to keep the questions alive across a semester

A term is a series of lectures, not one. Keep a single question list per course and add to it after every lecture, while the recording is still fresh. Write the questions the same evening. The ones you write a month later will be about the exam, not about the lecture.

Review the new questions the next day, before the next lecture arrives. That pass takes ten minutes, and it is the one most people skip. Without it the list grows, the reviews stop, and a long list of untested questions is just another set of notes.

Never delete a question because you got it right twice. Move it to the end of the list and answer it once a month. Deleting is how a question list turns back into a document you read.

By the end of a term the list is shorter than you fear. Most lectures produce a handful of definitions, one or two numbers and a method. If your list has sixty questions for one lecture, the questions are too small.

What active recall cannot do

It cannot create understanding you never had. Answering questions about material you never followed produces fluent nonsense, and you will mark it wrong without knowing why.

It does not replace problem practice in mathematics, physics or languages. There the unit of study is a worked problem, and you recall by solving rather than by reciting.

It depends entirely on the questions. Bad questions — too broad, several answers, drawn from a paragraph you did not understand — waste the session. Fix the questions before you blame the method.

It does not repair a bad transcript either. If the machine heard the lecturer wrong, you will memorise the error with great efficiency.

When active recall is not enough

• You cannot state the main point of the lecture in one sentence.
• The questions are copied from the slides and only ask you to define terms.
• You answer by glancing at the notes halfway through the sentence.
• The material is a procedure and you have never worked it end to end.
• You tested yourself once, on the day of the lecture, and not since.
• The transcript is machine-made and its key terms are wrong.
• You stopped writing new questions because the old ones feel familiar.