Short answer: self-explanation means stopping as you study, after a line of text or a step in a worked example, and explaining to yourself why it is true or how it connects to what you already know. A meta-analysis of 64 research reports (Bisra and colleagues, 2018) found a moderate-to-large average benefit (g = .55). It works best with open prompts you answer in your own words, and it pairs well with self-testing, which has even stronger evidence.
What is self-explanation?
Bisra and colleagues define it as a process by which learners generate inferences about causal connections or conceptual relationships. In plain terms: you don't just read "the heart's left ventricle has a thicker wall", you ask yourself why, and answer: because it pumps blood around the whole body, so it has to generate higher pressure.
The classic demonstration is Chi and colleagues (1994). Eighth-grade students read a textbook passage on the human circulatory system. Fourteen were asked to explain each line to themselves after reading it; ten read the same text twice without prompts. The prompted group gained more from pre-test to post-test, and prompted students who generated many explanations understood the system more deeply than those who generated few. Reading the text a second time did less.
What does the research say?
The biggest summary is Bisra, Liu, Nesbit, Salimi and Winne (2018) in Educational Psychology Review. They found 69 effect sizes from 64 research reports in which learners were prompted to self-explain while studying or solving problems. The overall weighted average effect was g = .55, which they compare to interventions such as peer tutoring. A few details matter for how you use it:
- It beat being given an explanation. Self-explanation outperformed reading an instructional explanation (g = .35), though a large part of the benefit over doing nothing extra was also available from good explanations by a teacher.
- Multiple-choice prompts didn't work. When students chose an explanation from a list, the effect (g = .24) was not statistically different from zero, though only two studies used that format. Generating your own explanation is the active ingredient.
- Explaining your own state of mind didn't help. Metacognitive prompts such as "do I understand this?" were the one type without a detectable effect. Explain the content, not your feelings about it.
For balance: Dunlosky and colleagues' (2013) review of ten study techniques rated self-explanation as moderate utility, below practice testing and spaced practice, mainly because the evidence was more limited and it takes time. Treat it as a strong partner to testing, not a replacement.
Self-explanation prompts that work
Keep them open, short and about the content. Pick one or two and use them at every paragraph or step:
- Why is this true?
- Why does this step follow from the last one?
- How does this connect to what I learned last week?
- What would happen if this were different?
- What rule or principle is this an example of?
Answer out loud or in a few words in the margin. Don't write essays; the benefit comes from the thinking, not the writing.
How to use it on notes, examples and answers
- On worked examples (maths, physics, chemistry). Cover the next step. Ask why it follows, predict it, then uncover. Self-explanation was originally studied on worked examples, and it suits material with a clear logic. Our guide to studying maths has more on worked examples.
- On textbook paragraphs. After each paragraph, explain to yourself why the main claim is true and how it links to the previous one, the way Chi's students did.
- On practice questions you get wrong. When you check the answer, don't just note the correct option. Explain why it is right and why your choice was wrong. This turns every mistake into a short self-explanation exercise.
- On practice questions you get right. Spend ten seconds explaining why. It catches lucky guesses.
- Then test without the notes. A day or two later, answer questions from memory. Explanation helps you understand; retrieval helps you keep it.
Self-explanation vs the Feynman technique and others
These techniques overlap, and that's fine; the differences are about when and to whom you explain.
- Self-explanation happens during study, line by line or step by step, to yourself.
- The Feynman technique happens after study: you explain a whole topic simply, as if to a beginner, to find the gaps.
- Elaborative interrogation focuses on "why" questions about facts, and is closely related.
- The protégé effect is about explaining to someone else, real or imagined.
If you only have time for one habit, make it testing yourself. If you have time for two, add self-explanation when checking your answers.
Checking answers is the natural moment to self-explain, and it helps to have the source in front of you. Ace turns your own notes, slides or PDFs into practice questions written from that material only, and under each answer it shows the sentence from your file that backs it, so you can ask "why is that the answer?" with the evidence right there.
If a whole topic has stopped making sense, our guide to studying for a test you don't understand shows how to find the gap before you start explaining.
Frequently asked questions
What is an example of self-explanation?
Reading "enzymes denature at high temperatures" and saying to yourself: "because heat breaks the bonds holding the active site's shape, so the substrate no longer fits." You are generating the reason, not just restating the fact.
Is self-explanation better than rereading?
Yes, in the research. In Chi and colleagues (1994), students prompted to self-explain gained more than students who read the same text twice.
Does self-explanation take longer?
Usually a little. Bisra and colleagues (2018) found the benefit held in studies where self-explanation took a similar amount of time to the comparison, which suggests it is time-efficient.
Can I self-explain using AI?
Write your explanation first, then compare it with an explanation from your notes or textbook. Reading an AI explanation instead of generating your own skips the part that does the work.
Sources
- Bisra, K., Liu, Q., Nesbit, J. C., Salimi, F., & Winne, P. H. (2018). Inducing self-explanation: A meta-analysis. Educational Psychology Review, 30(3), 703–725.
- Chi, M. T. H., de Leeuw, N., Chiu, M.-H., & LaVancher, C. (1994). Eliciting self-explanations improves understanding. Cognitive Science, 18(3), 439–477.
- Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students' learning with effective learning techniques. Psychological Science in the Public Interest, 14(1), 4–58.



