Back to blog

How to Know If You Really Know It: Judging Your Own Learning

Students are often confident about the wrong material. Why rereading feels like learning, the delayed self-test that fixes your judgement, and how to use confidence ratings while studying.

Most students decide what to study next by asking themselves "do I know this?" and the answer is often wrong. Material you just reread feels familiar, and familiarity feels like knowing, but it is a poor predictor of what you can recall on an exam days later. Researchers call these self assessments judgments of learning, and they become much more accurate with one change: judge yourself after a delay, by trying to recall the material without looking, instead of right after reading it. That delayed self-test is also good practice, so it improves both your judgement and your memory.

(This explainer is from the Nebulearn blog. Nebulearn makes a flashcard app, and flashcards happen to be one of the simplest ways to make these judgements honest.)

The evidence that we misjudge

StudyWhat happenedWhat it means
Roediger and Karpicke (2006)Students who reread a passage predicted they would remember more than students who had practiced recalling it. A week later, the recall group remembered substantially more.The method that feels better can be the one that works worse
Kornell and Bjork (2008)Students learning painters' styles thought blocked study (one painter at a time) worked better than interleaved study. Their test results showed the opposite.Smooth, easy study feels effective even when it is not
Nelson and Dunlosky (1991)People judged how well they would remember word pairs, either immediately after studying or after a delay. Delayed judgements were far more accurate.Wait before judging yourself, and test rather than reread
Koriat (1997) and later workPeople base judgements on cues like how fluently material reads, which often do not track memoryEase of reading is not evidence of learning

The pattern has a name in the literature: the illusion of competence, or the fluency illusion. Material in front of you, or material you just saw, is easy to process. Your brain reads that ease as "I know this."

Why immediate judgements fail

Right after you read a page, the information is still in short term memory. If you ask "will I remember this?", you check, find it, and say yes. But the question you actually care about is whether you can retrieve it in a week, from long term memory, without the page. An immediate check cannot answer that, because it is testing the wrong memory.

A delayed check can. If you wait even a few minutes, work on something else in between, and then try to recall the material without looking, you are testing something much closer to exam conditions. That is the core of the delayed judgment of learning effect that Nelson and Dunlosky found.

A practical method: predict, then test

StepWhat to doWhy
1. Study a sectionRead, take notes, work examplesNormal learning
2. Move onStudy something else for at least 10 to 15 minutes, or come back the next dayClears short term memory
3. PredictFor each key idea, rate how sure you are you can recall it: sure, unsure, or no ideaRecords your judgement before you check
4. TestClose everything and recall each idea, in writing or out loudThe real test
5. CompareCheck against your notes, and mark where your prediction was wrongShows how your judgement is calibrated
6. ReallocateSpend the next session on what you got wrong, especially "sure" items you missedTime goes where it is needed

Step 5 is where people learn the most about themselves. Most students find they are overconfident about material they reread several times, and sometimes underconfident about material they struggled with but actually learned.

Signs you are judging by familiarity

ThoughtWhat is usually going on
"I've seen this a lot, so I know it"Repetition of reading creates familiarity, not recall
"This makes total sense when I read it"Understanding while reading is not the same as producing it later
"I recognized the answer as soon as I flipped the card"Recognition, not recall. Grade it as a miss.
"I'll remember this, it's easy"Easy material is often skipped and then forgotten
"I know this, I just can't phrase it"Usually means partial knowledge, worth reviewing

The third one is the most common flashcard trap. You read the front, think vaguely, flip, see the answer and think "yes, that." That was recognition. Unless you produced the answer before flipping, it was not a correct recall. Passive vs active flashcard reviews and stop hitting Easy on flashcards cover how this distorts your schedule.

Flashcards as calibration tools

Flashcards make honest judgement easier because they force the test before the answer. A few habits make them more accurate:

  • Commit to an answer before you flip, out loud or in your head. A fuzzy "I sort of know" is not an answer.
  • Grade by what you produced, not by how the answer looked. If you only got half, it is not Good.
  • Use the Hard button for slow or effortful recalls. That is useful information for the scheduler.
  • Watch your own stats. If you mark almost everything Good but still fail practice exams, your grading is too generous.

Spaced repetition schedulers also do some of the judging for you. FSRS, the default in Nebulearn, estimates how likely you are to recall each card from your actual review history, not from how familiar the card feels, and brings it back when that probability drops. What is FSRS explains how it models this.

Using practice exams to calibrate

Practice exams are the best calibration check you have, because they test the way the real exam does. After each one:

  1. Before marking it, predict your score.
  2. Mark it, and compare.
  3. For each wrong answer, note whether you had felt confident about it.
  4. Turn confident errors into cards first. Those are the gaps your judgement is hiding from you.

If your predictions are consistently higher than your scores, assume you are overconfident and add more self testing to your plan. Post exam review flashcard strategy covers how to turn an exam into cards.

Where this goes wrong

  • Testing too soon. A self test right after reading still measures short term memory. Wait, or come back tomorrow.
  • Testing with the answer visible. Covering half the page and glancing is not a test.
  • Judging a whole topic at once. "I know cell biology" is too broad. Judge specific ideas.
  • Never acting on the result. Calibration only helps if you change what you study next.

Where to read more

Make It Stick by Peter Brown, Henry Roediger and Mark McDaniel has a full chapter on illusions of knowing. John Dunlosky and Janet Metcalfe's textbook Metacognition covers the research in depth. Retrieval strength vs storage strength explains the theory behind why easy feels like learning when it is not.

Test, then trust

Your sense of "I know this" is shaped by familiarity and fluency, which do not track what you can recall later. So do not judge right after reading. Wait, predict, test yourself without looking, and compare. Grade flashcards by what you actually produced, treat recognition as a miss, and turn confident errors from practice exams into cards first.

MetacognitionLearning ScienceSelf TestingStudy Techniques