How to Use the Feynman Technique When You Study

Close the book, explain the topic out loud in plain language as if to someone who has never met it, and notice exactly where you stall. That stall is the gap. Go back, fix only that, and explain again.

Four steps, and the one people skip is the first.

The four steps

1. Close everything. This is not a formality. Explaining with the notes open is reading aloud, and reading aloud is one of the least effective things you can do with study time. The whole method depends on producing the explanation from memory.

2. Explain it in plain words. Out loud, or written, to an imagined person who does not know the subject. No jargon unless you also define the jargon. If you find yourself saying a technical term as though it were an explanation, that is a gap wearing a costume.

3. Find where you stalled. You will hit a point where the sentence will not finish, or where you said something vague to move past it. Mark it. That specific point is what you did not understand, and it is usually much narrower than “I do not get this topic”.

4. Go back for that piece only. Not the whole chapter. Open the notes, resolve the one thing, close them, and explain again from the start.

Why it works

The method is a retrieval exercise wearing a teaching costume. You are producing material from memory under a constraint, and the constraint exposes the difference between recognising something and being able to generate it.

That distinction is the whole game. Reading a paragraph three times produces a strong feeling of familiarity and very little ability to reproduce it. Practice testing and distributed practice came out as the two highest utility strategies in Dunlosky and colleagues’ 2013 review of ten study techniques, with rereading and highlighting near the bottom. The Feynman technique is practice testing that you can run without anyone writing questions for you.

The plain language requirement does a second job. Technical vocabulary lets you assemble a sentence that sounds correct without understanding what it describes. Strip the vocabulary and the sentence either survives or it does not.

Where people go wrong

Mistake	Fix
Explaining with the notes open	Close them. Produce it from memory or the exercise does nothing
Using jargon as the explanation	Every technical term must be defined in ordinary words
Rereading the whole chapter when you stall	Go back for the one specific piece, then close and restart
Doing it once and declaring it done	Repeat on a different day. The gaps move
Only using it on topics you already like	The subject you avoid is the one with the most gaps in it

What a real session looks like

Pick one concept. Not a lecture, not a module. One mechanism, one process, one argument.

Set a timer for five minutes and talk. Out loud matters more than it sounds: your inner monologue skips over gaps silently, and your speaking voice does not, because a sentence has to finish.

When you stall, write the stall down in words. “I do not know why the second step needs the first one to have happened.” That phrasing is now a question you can answer, which is more useful than a vague sense that a section was confusing.

Resolve it, close the notes, and go again from the beginning. The second run is usually noticeably faster, and the place you stall will have moved further along.

Two or three rounds is a session. It is more tiring than rereading, which is the point.

Adapting it for different subjects

Processes and mechanisms. The method’s natural home. Explain what happens, then explain why each step needs the one before it.

Maths and problem subjects. Explain the method rather than the answer. Say why this approach applies to this problem and not another one. Working through the arithmetic is a different exercise.

Essay subjects. Explain the argument and then the strongest objection to it. If you can only produce the argument, you have half of what the marker wants.

Memorisation heavy material. The Feynman technique is a poor fit for arbitrary facts. Drug doses and vocabulary want spaced retrieval, not explanation. Use it for the parts that have structure and reasons.

Combining it with your notes

The natural rhythm is: attend the lecture, capture the cues, then within a day try to explain the main idea closed book before you review anything. Do the explanation first and the reading second. Reversing that turns the whole session back into rereading, and it is a very easy reversal to make without noticing.

Keep a running list of the stalls. After a fortnight that list is the most accurate map of what you do not know that you will ever have, and it was produced by you rather than guessed by a revision guide.

Where a tool fits

The awkward part of the method is the moment after you stall. You need a targeted answer to a narrow question, from your own material, without opening the whole chapter and losing the thread.

That is what StudyLabAI’s Feynman technique feature is for. Upload the lecture, explain the concept, and get questions and a check against that specific material rather than a general explainer on the topic. The prompting and the checking are handled. The explaining is the part that has to be yours, because the explaining is the studying.

Common questions

Do I have to say it out loud?
Out loud or written. Thinking it through in your head lets you skip gaps without noticing, because an unfinished thought still feels complete.

Who am I supposed to be explaining to?
Someone intelligent who has never studied the subject. A friend on a different course is the usual standard. Not a child, which pushes people into oversimplifying, and not your lecturer, which lets you hide behind jargon.

How is this different from just testing myself?
It is a form of self testing, with a constraint that catches a specific failure: knowing the words for something without knowing the thing. Ordinary questions can be answered with a memorised phrase. An explanation cannot.

How long should a session be?
Five to fifteen minutes per concept, two or three rounds. It is demanding, and doing it for an hour usually means you drifted into rereading somewhere in the middle.

Does it work for exams that are mostly recall?
Partly. Use it for anything with a mechanism or a reason behind it, and use spaced retrieval for the arbitrary facts. Most courses need both.