Why 'I Get It' Isn't the Same as 'I Remember It'

The short answer: "I get it" is the brain congratulating itself for recognising, not for knowing

There is a thing students do at the end of an evening of studying that feels like the win condition: they look up from the textbook and think "yes, I get this now." Robert Bjork's research at UCLA on judgments of learning (source) is what makes this finding survive contact with reality: that feeling of "I get it" is almost completely uncorrelated with whether you can actually retrieve the material a day later. The fluency you feel while reading is the brain reading easy text — fluently — and reading that ease as understanding. The fluency is real; the understanding is borrowed from the page that's still in front of you.

Why this is the single most expensive learning bug

Most students do not undertest themselves out of laziness; they undertest because the fluency illusion convinces them they don't need to. The reasoning is: "I just read this and I clearly understand it. Why would I test myself? I'd ace it." The next morning's blank reveals that the test would in fact not have been aced. By then the study session is over and the false sense of understanding has been booked as learning that didn't happen. Compounded over a semester, this is the difference between feeling like a good student and actually being one.

How to tell fluency from knowledge — five tests

• Close the book and explain. If the explanation falls apart without the source visible, you had fluency, not knowledge. This is the Feynman test, the spaced-retrieval test, and the fluency-illusion test all wrapped together — it cleanly separates familiarity (recognising what's true) from recall (producing what's true).

• Predict the test result, then take it. Bjork's lab has used this as a diagnostic for years. Write down what score you'd expect on a small recall test; then take the test. The gap is the calibration error. Most readers find the gap is much larger than they thought — that gap is the fluency illusion measured directly.

• Wait twenty-four hours. Fluency decays fast; knowledge decays slowly. The test you can pass at 9pm the night you studied is not the test that matters. The test that matters is the one you can pass when you've slept, done other things, and come back cold. Schedule the recall attempt for the next morning, not the same session.

• Use deliberately hard practice. Interleave problem types, mix in old material, prefer problems slightly above your current level. Bjork calls these "desirable difficulties" — they make the session feel less fluent, which means less false confidence and more actual learning. The discomfort during practice is the signal you're producing knowledge rather than fluency.

• Teach it out loud, with no notes. Teaching is the strongest fluency-illusion solvent available. The act of having to choose words, sequence ideas, anticipate the listener's confusion exposes the holes that the silent re-read never finds. The Feynman technique exists because this is one of the highest-yield diagnostic moves the literature has documented.

Why this pays double for ADHD

ADHD readers tend to find the fluency illusion especially strong because the felt experience during study sessions is more sensitive — when something clicks, the click is loud, and the click feels conclusive. The next morning's blank then registers as a personal failure ("I had it last night") rather than as the predictable behaviour of the fluency illusion. Knowing this in advance changes how you read the blank — not as evidence of weak character but as evidence that the wrong thing was being measured the night before. The technique stays the same; the emotional load drops.

Where it fails (and the repair)

• Trusting recognition tests. Multiple-choice questions, glancing at flashcards, scanning bullet summaries — all of these test recognition, not recall. They produce false-positive fluency readings. The fix is to insist on free-recall: produce the answer with nothing on the page, then check.

• Calling a re-read "reviewing". Re-reading produces the strongest possible fluency without producing the strongest possible knowledge. The cure is to call it what it is — re-reading — and not let it count as the review pass. Replace one re-read per session with a free-recall attempt and the learning curve changes shape immediately.

• Stopping as soon as it feels easy. The moment a concept feels easy is the moment the fluency illusion is strongest. That's the moment to test, not to stop. The hardest psychological move is to test yourself right when you feel you don't need to, and that's also where the highest learning value lives.

FAQ

If I feel I understand it but can't recall it later, did I understand it or not?

You experienced something real — the sensation of fluency-with-text-in-front-of-you is not lying about how easy the text was to process. What it's lying about is whether that ease will transfer to a test without the text. Both states are real; you're just calling them by the same word. Disambiguate: "this is fluent in my head right now" vs "I can produce this cold". The first is real and short-lived; the second is the one schools and life measure.

Does this apply to skills as much as facts?

Yes, with the appropriate translation. The motor-skill analogue of fluency is "I can do this when the conditions are perfect". The knowledge equivalent is "I can do this when conditions vary". Pianists, surgeons, programmers all face the same fluency-illusion: rehearsing under easy conditions feels productive but doesn't transfer. Add variation — sloppy conditions, time pressure, different problem shapes — and you replace fluency with transfer.

How often do I need to retest the same material?

Use the spacing schedule from the spaced-repetition piece: roughly 10-20% of the desired retention interval. For a week, retest every day or two. For a semester, retest every couple of weeks. The compounding of spaced retrieval is the strongest known antidote to the fluency illusion, because each retest forces a recall under colder conditions than the last.

Why do textbooks not just tell students this?

Some do, but most are organised around the assumption that students will use them in a particular way (read, summarise, re-read, take a test). The fluency illusion is a finding from psychology that has been slow to migrate into standard study habits. Most students who do well anyway either stumbled into spaced retrieval naturally or had a teacher who insisted on it. The information is now widely available; using it is the question.

Is there a quick gut check during study?

Yes: close the book mid-paragraph and say what the next paragraph is going to argue. If you can, you're producing knowledge; if you can't, you've been recognising what's on the page and miscoding the recognition as learning. Do this two or three times in a study session and the fluency illusion stops fooling you in real time, not just retrospectively.