Skip to content

Study techniques and their evidence

Elaborative interrogation and self-explanation: explaining things to yourself

How learning works · Published · Evidence: Moderate · About this guide

Elaborative interrogation means asking yourself why a fact is true; self-explanation means putting into words how each new step or idea follows from what you know. Both tend to help: in a 2018 meta-analysis, learners prompted to explain did better by about half a standard deviation on average [1][2]. Most tests came within minutes, explaining takes extra time, and in a six-month comparison testing yourself did better [3].

Nodding along while you read is easy. Stopping to ask "why is that so?" and answering in your own words takes more effort. For many kinds of material, studies suggest the effort pays off. This article covers what they found, what it costs in time, and where the evidence is still thin.

What does "explaining to yourself" mean?

Researchers have studied two closely related habits. The first is called elaborative interrogation: when you meet a fact, you stop and ask yourself why it is true. Most studies used a sharper form of the question: why would this be true of this thing, and not of something similar [1]? The second is self-explanation: you put into words how a new piece of information connects to what you already know, or why each step of a worked example follows from the one before [1][2].

For "why?" questions, the prevailing account is that answering ties a new fact to what you already know, while the "not of something similar" part helps you tell related facts apart. Self-explanation may work in a similar way, but its studies prompt explanations in too many different ways for the reviewers to say why it works [1]. A large 2013 review of ten common study techniques rated both as having "moderate utility". There is real evidence behind them, but they have been tested less widely than practice testing and spaced practice, the only two techniques the review rated as high utility [1].

What did the early studies find?

In an early series of three experiments, university students read simple sentences such as "The hungry man got into the car" [4]. One group just read the sentences. A second group was given a reason with each one ("…to go to the restaurant"). A third group had to answer a question for themselves: why did that particular man do that? On a later test of which man did what, the group that produced its own reasons remembered roughly twice as much: about 72 % correct, against about 37 % in each of the other two groups, averaged across the experiments [1]. Being handed a reason helped much less, and only under some conditions; coming up with it yourself gave much larger gains [4].

A second classic study looked at understanding rather than memorising lists [5]. Fourteen eighth-grade students read a passage on the human circulatory system and were asked to explain to themselves what each line meant. Ten others read the same passage twice. The students asked to explain improved more between a test before reading and a test after it [5], although explaining also took them about twice as long [1]. Among them, those who produced many explanations all ended up with a correct picture of how the circulatory system works; many of the other students did not [5].

How large is the effect across many studies?

A 2018 meta-analysis combined 69 results from 64 reports, covering nearly 6,000 learners from primary school to professional training [2]. On average, learners prompted to explain did better by about half a standard deviation (g = 0.55): a moderate effect, not a dramatic one. It appeared when people studied texts, solved problems or worked through examples, in science, maths and other subjects, and at every level of schooling the authors looked at. Results varied a good deal between studies, so the average hides real differences [2].

The gain was larger against learners who got no extra explanation (g = 0.67), and smaller but still there against learners who were handed a ready-made explanation of the same points (g = 0.35, in only six comparisons) [2]. So working an explanation out yourself may add something beyond reading one, but the two figures come from different studies, and the second from few.

Why is it rated only "moderate"?

The reviews give three main reasons.

It has mostly been tested within minutes. Almost every study tested learners straight after the learning session. The few longer-term results are encouraging, but many of them had an earlier test or extra study in between, and the one-week test in the circulatory-system study was open-book [1]. How long the advantage lasts is still largely an open question.

It takes time. In the few self-explanation studies that reported how long learners took at their own pace, explaining usually took 30–100 % longer than the comparison activity; in the circulatory-system study, the explaining group spent about 125 minutes against 66 for the rereaders. "Why?" questions looked cheaper: in one self-paced study they took 32 minutes against 28 for reading alone [1]. In the meta-analysis, the five comparisons where both groups spent about the same time still found a benefit (g = 0.41), smaller than where explaining took longer (g = 0.72), although that difference was not statistically reliable [2]. In one study, children who explained while solving maths problems did about as well as children who simply solved twice as many problems [1].

It depends on what you already know. Asking "why?" seems to help most when you already know something about the topic. In one study, Canadian and German students learned facts about Canadian provinces and German states, and the benefit was about twice as large for facts about their own country [1]. For people new to a subject, the benefit is less certain. "Why?" questions also work most clearly with separate facts; for a complex system, it is not clear what to ask "why" about [1].

How does it compare with testing yourself?

Explaining is not the same as recalling. In a well-known pair of experiments, students read a science text and then either drew a concept map of it with the text in front of them (an elaborative activity, though not the same as explaining) or put the text away, wrote down everything they could remember, reread it and recalled again [6]. Study time was matched. A week later the recall group did clearly better, including on questions that required connecting ideas, and in the second experiment even when the final test asked them to draw a concept map. Most students had expected concept mapping to work at least as well [6]. The paper drew a published critique and a reply, and a 2021 follow-up from the same lab again found recall ahead of concept mapping after a week [7]. A preregistered 2023 replication by another team found the same advantage at first, but it disappeared when a concept-mapping group, in the same total time, also spent its last five minutes memorising its map [8]; the article on retrieval practice describes that study.

Explaining itself has been set against testing less often. In one study, 47 first-year medical students learned four clinical topics in four weekly sessions, each topic by a randomly assigned activity: tests, tests plus explaining, explaining while studying a review sheet, or just studying the sheet [3]. Six months later, on a test of recalling the material and applying it to clinical situations, they scored on average about 36 %, 40 %, 29 % and 20 % respectively. Both kinds of testing beat explaining with the sheet, and explaining beat plain study. Results varied by topic, and the small gap between tests with and without explaining leaves open whether the combination beats testing alone [3].

With so few studies, treat these results as strong hints rather than a final verdict: explaining looks more like a complement to recall than a replacement [3].

Is this the Feynman technique?

The studies here tested self-explanation and elaborative interrogation, not the "Feynman technique" as a package. The evidence on explaining ideas in your own words is in the sections above; the evidence on recalling them from memory is in the article on retrieval practice.

How can you try it?

  • Ask the sharper question. At a fact you want to learn: why is this true, and why of this and not of something similar [1]?
  • Explain, do not just reword. When working through an example, say why each step follows from the last. Rewording a sentence is not explaining it; in one study, students with weak algebra skills mostly paraphrased when asked to explain [1].
  • Try before you look. In one study, learners who could look up a ready-made explanation seemed to make less effort of their own, and the benefit of being asked to explain showed up only among those who could not look it up [1][2].
  • Ask often, in small steps. In longer texts, one "why?" prompt every page or two weakened the effect in one study and reversed it in another [1].
  • Check your explanation afterwards. Compare it with the source. An explanation can feel convincing and still be wrong.

What the research does not say

  • That explaining replaces testing yourself. In the study that compared the two over six months, testing did better [3]. With study time matched, recall also beat concept mapping, another elaborative activity [6][7], although in one replication that advantage disappeared once the concept mappers also had time to memorise their maps [8]; and the 2013 review rated both explaining techniques below practice testing and spaced practice [1].
  • That one session of explaining lasts for months. Most tests came within minutes, and the few longer-term results were often mixed with other practice [1]. The 2018 meta-analysis did not examine how long the effect lasts [2]. The six-month result above came after four weekly sessions [3].
  • That explaining more makes you learn more. In the circulatory-system study, the "high" and "low" explainers were not assigned; they were sorted afterwards by how much they explained. People who already know more may both explain more and learn more, so the link does not show cause [5][1].
  • That it works the same for everyone. The benefit of "why?" questions is less certain for beginners in a subject, and for self-explanation the role of prior knowledge has not been tested systematically [1].
  • That it costs no extra time. Where time was reported, explaining usually took longer than the comparison activity, and equal-time studies are few [1][2].
  • That lab numbers carry over to a course. The 72 % against 37 % came from lists of simple sentences. In a university biology course, students who answered "why is this true?" prompts while reading a textbook chapter scored 76 % on a true/false test taken straight afterwards, against 69 % without prompts; on such a test, guessing alone would score about half [1].
  • That an unchecked wrong explanation is harmless. These studies measured what happens when people are prompted to explain. They do not tell us what a confident but wrong explanation costs when nobody corrects it.

Sources

Every source the guide cites

  1. [1] Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students' learning with effective learning techniques: Promising directions from cognitive and educational psychology. Psychological Science in the Public Interest, 14(1), 4–58. review · doi.org/10.1177/1529100612453266
  2. [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. meta-analysis · doi.org/10.1007/s10648-018-9434-x
  3. [3] Larsen, D. P., Butler, A. C., & Roediger, H. L., III (2013). Comparative effects of test-enhanced learning and self-explanation on long-term retention. Medical Education, 47(7), 674–682. experiment · doi.org/10.1111/medu.12141
  4. [4] Pressley, M., McDaniel, M. A., Turnure, J. E., Wood, E., & Ahmad, M. (1987). Generation and precision of elaboration: Effects on intentional and incidental learning. Journal of Experimental Psychology: Learning, Memory, and Cognition, 13(2), 291–300. experiment · doi.org/10.1037/0278-7393.13.2.291
  5. [5] Chi, M. T. H., de Leeuw, N., Chiu, M.-H., & LaVancher, C. (1994). Eliciting self-explanations improves understanding. Cognitive Science, 18(3), 439–477. experiment · doi.org/10.1207/s15516709cog1803_3
  6. [6] Karpicke, J. D., & Blunt, J. R. (2011). Retrieval practice produces more learning than elaborative studying with concept mapping. Science, 331(6018), 772–775. experiment · doi.org/10.1126/science.1199327
  7. [7] O'Day, G. M., & Karpicke, J. D. (2021). Comparing and combining retrieval practice and concept mapping. Journal of Educational Psychology, 113(5), 986–997. experiment · doi.org/10.1037/edu0000486
  8. [8] Mayrhofer, R., Kuhbandner, C., & Frischholz, K. (2023). Re-examining the testing effect as a learning strategy: the advantage of retrieval practice over concept mapping as a methodological artifact. Frontiers in Psychology, 14, Article 1258359. experiment · doi.org/10.3389/fpsyg.2023.1258359

This guide is general information about learning research for students, not medical advice.