Does the Feynman technique work?

What the technique is

Named after the physicist Richard Feynman, the technique has a few steps: pick an idea, explain it as if to someone who knows nothing about it, notice where your explanation breaks down or falls back on jargon, go back to the source, and try again. Researchers study the same thing under other names: self-explanation, elaborative interrogation (asking yourself why something is true) and learning by teaching.

The illusion it exposes

People reliably overestimate how well they understand how things work. In studies of the illusion of explanatory depth, people rated their understanding of everyday mechanisms, such as how a zip or a toilet works, then tried to explain them step by step, and their ratings dropped [1]. The illusion is stronger for mechanisms than for facts. Trying to explain is one of the few things that reliably punctures it.

What the studies found

Self-explanation helps. A meta-analysis of 69 effect sizes from 5,917 learners found that prompting people to explain what they were studying improved learning by about half a standard deviation (g = 0.55) compared with no prompts [2]. It even beat being given a good explanation (g = 0.35): producing the explanation yourself is part of the benefit. Prompts that simply asked learners to pick an explanation from options had no clear effect.

Teaching helps, especially out loud. Students who prepared to teach learned more than those who studied normally, and actually teaching helped more still (g = 0.35 and 0.56) [3]. In a meta-analysis of explaining to a fictitious student, explaining aloud helped (g = 0.34), while writing the same explanation did not (g = −0.07) [4]. A related experiment found the same pattern: explaining on video helped (d = 0.43), explaining in writing much less (0.19, not significant) [5].

Elaboration grows with knowledge. Dunlosky and colleagues rated elaborative interrogation and self-explanation as moderate utility: well established, but with benefits that depend on the material and on what the learner already knows [6].

When explaining doesn’t help

• When there is nothing to explain. Self-explanation is of limited value where no inference is possible: arbitrary facts, lists, definitions [7]. In one study it even hindered learning categories full of exceptions, because explaining pushes you to find a pattern that isn’t there.
• When it replaces checking. An explanation that sounds fluent can still be wrong. Explaining works best followed by checking against the source.
• When you already understand it. Generative tasks help most while you are still working something out; once you understand it, recalling it on a schedule does more [8].

How to use it well

1. Choose a mechanism, not a fact. “Why do the kidneys hold on to fluid in heart failure?” is worth explaining. “What is the normal sodium level?” is better as a card.
2. Explain aloud, in order, without your notes. Say each step and why it leads to the next.
3. Mark the gaps. Wherever you hesitate, wave your hands, or use a term you couldn’t define, write it down.
4. Go back to the source for just those gaps. Then explain again.
5. Apply it to a new case. The real test of understanding is using the idea on a situation you haven’t seen. Then come back to it in a few days. See retrieval practice.

References

1. Rozenblit, L., & Keil, F. (2002). The misunderstood limits of folk science: An illusion of explanatory depth. Cognitive Science, 26(5), 521–562. (Abstract.) onlinelibrary.wiley.com/doi/abs/10.1207/s15516709cog2605_1
2. 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. link.springer.com/article/10.1007/s10648-018-9434-x
3. Kobayashi, K. (2019). Learning by preparing-to-teach and teaching: A meta-analysis. Japanese Psychological Research, 61(3), 192–203. (Abstract.) onlinelibrary.wiley.com/doi/10.1111/jpr.12221
4. Lachner, A., Jacob, L., & Hoogerheide, V. (2021). Learning by writing explanations: Is explaining to a fictitious student more effective than self-explaining? Learning and Instruction, 74, 101438. www.sciencedirect.com/science/article/abs/pii/S0959475220307337
5. Hoogerheide, V., Deijkers, L., Loyens, S. M. M., Heijltjes, A., & van Gog, T. (2016). Gaining from explaining: Learning improves from explaining to fictitious others on video, not from writing to them. Contemporary Educational Psychology, 44–45, 95–106. www.sciencedirect.com/science/article/abs/pii/S0361476X16000102
6. 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. doi.org/10.1177/1529100612453266
7. Wylie, R., & Chi, M. T. H. (2014). The self-explanation principle in multimedia learning. In R. E. Mayer (Ed.), The Cambridge handbook of multimedia learning (2nd ed., pp. 413–432). Cambridge University Press. education.asu.edu/sites/g/files/litvpz656/files/lcl/wylie_chi_selfexplanation_0.pdf
8. Roelle, J., & Nückles, M. (2019). Generative learning versus retrieval practice in learning from text: The cohesion and elaboration of the text matters. Journal of Educational Psychology, 111(8), 1341–1361. eric.ed.gov/?id=EJ1232781