The classic experiment gives people a list of adjectives and asks a different question about each one.
Is it printed in capitals? Does it rhyme with another word? What does it mean? Does it describe you?
Then, unexpectedly, everyone is asked to recall the words. Recall climbs steadily across those four conditions, and the self-descriptive question wins by a wide margin. Judging whether a word applies to you produces better memory for it than judging its meaning, which already beats judging its sound or its appearance.
That is the self-reference effect, and it has replicated for decades across many variations.
Why it works
The standard explanation is that the self is an unusually rich and well-organised structure in memory. Relating something to it connects the new item to an enormous existing web of associations, which gives retrieval many more routes in.
It fits neatly on top of levels of processing: the deeper you process something, the better it is remembered, and processing against your own experience is about as deep as ordinary thinking goes. Self-reference is not a separate mechanism so much as the far end of that scale.
| What you do with the material | Depth | Retention |
|---|---|---|
| Copy it verbatim | Shallow. You process shape | Poor |
| Reword it | Moderate. You process meaning | Better |
| Explain it to someone | Deep | Good |
| Connect it to your own experience | Deepest | Best |
What this actually changes about note-taking
Most note-taking advice tells you to use your own words and stops there, as if it were a style preference. It is not. It is the difference between a shallow condition and a deep one in the strongest memory finding available.
The concrete practices:
Never copy a definition. Write what it means for something you do. A note reading "Loss aversion: losses feel roughly twice as bad as equivalent gains" is fine. A note reading "Loss aversion is why I could not cancel a subscription I stopped using, twice" is memorable, because it is anchored to you.
Add the personal hook line. One sentence per note: where did I meet this, what does it change for me, what does it contradict that I believed. This costs ten seconds and moves the note up two rows in the table above.
Write down the disagreement. Noticing that you do not believe something is intensely self-referential and produces unusually durable memory. It is also the seed of anything original you will ever write. See writing from your notes.
Use examples from your own life rather than the textbook's. The textbook's example is someone else's self-reference.
This is why atomic notes and evergreen notes insist that a note state a claim rather than name a topic: a claim is something you can agree or disagree with, and a topic is not.
The overlap with everything else that works
The self-reference effect explains part of several better-known techniques rather than competing with them.
The generation effect says material you produce is remembered better than material you receive. The Feynman technique works partly because explaining forces you to route the idea through your own understanding. Elaborative interrogation, asking why is this true, tends to pull in personal knowledge as evidence.
They are converging on one instruction: do something to the material that only you could have done.
The honest limits
Three caveats worth having.
It is not magic, it is deeper encoding. If the material genuinely has nothing to do with you, forcing a personal connection produces strained analogies that can interfere rather than help. Some things are just facts to be learned by spaced repetition.
The effect is measured on memory, not understanding. Relating quantum mechanics to your feelings will not teach you quantum mechanics. Self-reference is an encoding advantage, not a comprehension one.
Self-relevance can distort. The same machinery that makes personally relevant material memorable also makes it selectively remembered. Notes about a conflict at work will encode your side vividly and the other side thinly, and you will not notice. For anything you may need to reconstruct fairly, record what was said separately from what you concluded. See false memory and how to take interview notes.
The one-line version
If a note could have been written by anyone, it will be remembered by no one. Put yourself in it.
More in memory and focus, why writing helps memory, and how to take better notes.