The illusion of competence is the feeling of having learned something, formed while the material is still in front of you. It comes from judging your knowledge in a situation where the answer is present, when the test will ask for it with the answer absent. It is not the Dunning-Kruger effect, although the second result on Google says it is. The fix follows from the cause: move the judgment to a moment when the answer is gone.
What is the illusion of competence?
The term comes from memory research. Asher Koriat and Robert Bjork named it in a 2005 paper in the Journal of Experimental Psychology: Learning, Memory, and Cognition (31(2), 187-194). Their argument is that studying has a built-in mismatch. While you study, you rate how well you know something with the full information on the page. At test, the same information is missing, and you are asked to produce it.
People fail to discount what is on the page when they make that rating. So they come away with a sense of competence that the test then does not confirm.
A small illustration of the mechanism. Read the pair "ocean: tide" and ask yourself whether you will remember it. It feels obvious, because once you see both words, the link between them is plain. Tomorrow you get "ocean: ___" alone, and "wave", "salt" and "blue" all come up first. The link that felt strong only existed with both halves in view.
That is the whole illusion. Nothing about it is a character flaw or a lack of intelligence. It is a property of the situation, and it happens to careful, skilled learners too.
Is the illusion of competence the same as the Dunning-Kruger effect?
No, and the difference changes what you should do.
The Dunning-Kruger paper is Kruger and Dunning, 1999, in the Journal of Personality and Social Psychology (77(6), 1121-1134). Across four studies, people who scored in the bottom quarter on tests of humor, grammar and logic placed themselves around the 62nd percentile. Their actual scores put them in the 12th. The claim is about people who lack a skill also lacking the insight to see it.
The illusion of competence is a different finding in three ways:
- Who has it. Dunning-Kruger describes a gap between people. The illusion of competence hits everyone, strong students included, because it comes from the study situation.
- When it happens. Dunning-Kruger is about overall self-assessment. The illusion of competence forms in the moment of studying, item by item.
- What cures it. Dunning-Kruger shrinks when people get better at the skill. The illusion of competence shrinks when you change when and how you judge your own learning.
Treat them as one thing and you get advice like "be more humble". Humility does nothing about an answer that is sitting on the page while you rate yourself.
Why do podcasts and lectures produce it so reliably?
Because everything that makes spoken material pleasant also makes it feel learned.
Shana Carpenter and colleagues tested this directly in 2013 (Psychonomic Bulletin & Review 20(6), 1350-1356). Participants watched one of two short videos of an instructor explaining the same scientific concept. In one, the instructor stood upright, kept eye contact and spoke fluidly without notes. In the other, the instructor slumped, looked away and spoke haltingly from notes. People who saw the fluent version believed they had learned more, and rated the instructor as better prepared and more effective. They had not learned more. The amount learned did not differ significantly.
A podcast host is the fluent instructor, edited. Pauses are cut, the argument arrives in order, and the voice is confident. Every quality you would praise in a good episode is a quality that raises your sense of having learned it. The same goes for a beautifully drawn lesson, as our Imprint app review shows.
Then there is the mismatch Koriat and Bjork described. While you listen, the answer is always present, because someone is saying it. You never once sit in the test condition during an episode. Recognizing each sentence as it arrives feels like understanding, and this is the audio version of the rereading trap described in how to remember what you read.
Lorena Barba, an engineering professor at George Washington University, put it in a sentence quoted by The Chronicle of Higher Education: think of an inspiring lecture that left you feeling as if you had heard something revelatory, only to realize later that you did not get anything from it.
Why does real learning feel worse?
The mirror image of Carpenter's result appeared in a 2019 study by Louis Deslauriers and colleagues in PNAS (116(39), 19251-19257). They compared passive lectures with active learning in a randomized setup, using identical course material. Students in the active class learned more but felt that they had learned less. The authors trace part of that gap to the extra cognitive effort that active learning demands.
So the signal most people use to judge a learning session runs backwards. Easy feels productive. Effortful feels like failure. If you choose your methods by how they feel, you will keep choosing the ones that teach you least.
How do you get rid of the illusion?
Koriat and Bjork followed up in 2006 in Memory & Cognition (34(5), 959-972) and tested what reduces it. Two things worked: experience with the test, and delaying the judgment until some time after study. Both have the same effect. They put you in the answer-absent condition before the real test does.
For an episode or a chapter, that turns into a short routine:
- Wait before you judge. Do not ask "did I get that?" as the episode ends. Ask the next morning, when nothing is playing. The night in between is not wasted time either: sleep keeps memories it expects you to need, as the studies in can you learn while sleeping show.
- Recall with the source closed. Say or write the main argument without the app open. Two or three sentences is enough to find the holes.
- Only then check. Open the episode or your notes and compare. What you missed is the real state of your knowledge.
- Take the test that exists. A few questions on the material, answered from memory, is the closest thing to what the 2006 study called test experience.
Explaining the material to another person is a strong version of step 2, because a listener who asks back finds gaps you would skip on your own. That is covered in more detail in the protégé effect. If you want to add structure to the testing part, active recall apps for podcasts compares the options.
None of this needs special software. A notebook and the discipline to close the app before you write work fine. What makes it hard is that every step feels less pleasant than just listening to the next episode, which is exactly what Deslauriers measured.
Where does Graspd fit?
Graspd is built around this problem, so read this section as an interested party talking. It is an audiobook and podcast companion that quizzes and recaps you on what you have heard, using only the transcript up to your current position so nothing later gets spoiled. Nothing stops you from opening the quiz the moment an episode ends, so the delay from the 2006 study is still yours to add: take it the next day. You drop marks while listening, and the companion uses them when you talk the episode through, the loop described in how to remember what you listen to.
Two honest limits. None of the studies above tested an app, so the evidence supports the method, not any particular product. And Graspd cannot import Audible audiobooks, because they are protected by DRM. Podcasts work, and audiobooks work when you have the actual file.
What do the popular explanations get wrong?
Two of the three top results for this term have problems you can check yourself.
The most detailed ranking explainer, from Coursera, states that the illusion of competence is "also known as the Dunning-Kruger effect" and cites the 1999 Kruger and Dunning paper as its source. That paper measures a different thing, and the research that coined the term is about the study situation, not about unskilled people. The same article then spends a section on the four stages of competence, a skill-training model that explains neither finding.
The first result is a university news item that excerpts one paragraph from The Chronicle of Higher Education and links out. It has the best single example of the three, the inspiring lecture, but no mechanism and no remedy. The third is a discussion thread about students using AI. None of them mentions that the fix is to change when you judge your learning, which is the finding with the most practical use.
FAQ
What is an example of the illusion of competence?
Finishing a podcast episode on a complex topic, feeling that you understood every part, and then being unable to explain its main argument the next day. You followed each sentence while the host was saying it, which felt like knowing it. The test, explaining it without the episode playing, came later and showed the gap.
Is the illusion of competence the same as the Dunning-Kruger effect?
No. Dunning-Kruger describes people with low skill overestimating their ability compared to others. The illusion of competence happens to everyone during study, including skilled learners, because you judge your knowledge while the answer is still in front of you. The remedies differ too: the first needs more skill, the second a delayed self-test.
Who coined the term illusion of competence?
In memory research, Asher Koriat and Robert Bjork used it as the title of a 2005 paper on how people judge their own learning during study. Karpicke, Butler and Roediger applied it in 2009 to students who reread instead of testing themselves. The phrase was later popularized for general audiences through learning courses and study guides.
Why do I feel like I learned more from a good speaker?
Because fluency raises your sense of learning without raising what you actually learn. In a 2013 study, people who watched a smooth, confident instructor believed they had learned more than people who watched a halting one explain the same concept. Test results showed no significant difference. Polished podcasts create the same effect.
How can I tell if I really understood something?
Wait a day, close the source, and try to explain the main point in two or three sentences. Then check against the original. What you could not produce without the source was not learned yet, however clear it felt at the time. Delaying the check is what makes it honest.
