Francis Bacon knew the testing effect four centuries before psychologists proved it
The best way to learn something was written down in 1620 and then forgotten by most schools anyway.
In 1620, Francis Bacon observed that reading a text twenty times over teaches it less securely than reading it ten times while repeatedly stopping to recite it from memory, checking the page only when memory fails. He wasn't running an experiment. He was just noticing something that would take psychology nearly three more centuries to formally confirm.
In 1909, researcher Edwina Abbott published what's usually credited as the first controlled study of what's now called the testing effect: material that gets retrieved from memory during study sticks better than material that only gets re-read, even when both groups spend identical time on the page. The gap isn't small, and it isn't about the material being easier — if anything, the harder the retrieval attempt, the stronger the later memory, provided the retrieval eventually succeeds.
Schools have mostly built their systems around the opposite intuition. Studying looks like re-reading notes, highlighting a textbook, watching a review video — all recognition-based, all comfortable, all weaker than closing the book and trying to recall the answer cold. Testing gets treated as the thing that happens after learning, to measure it, rather than the thing that produces the learning in the first place.
Bacon had it in one sentence. It took until 1909 for anyone to write it down as data.