What Is Elaborative Interrogation? Getting Learners to Explain Why

Most online lessons tell learners things. A fact, a rule, a step in a process, then the next screen. Learners nod along, and the course moves on. Elaborative interrogation is the habit of stopping at the important facts and asking the learner one question: why?

Elaborative interrogation (asking learners to explain why a stated fact or idea is true) is one of the better-supported learning strategies, alongside its close relative, self-explanation. The evidence is real but more modest than for retrieval practice, and the popular guides to it, mostly written for classroom teachers and students revising for exams, skip the parts that matter when you build a self-paced course: how much time the explanations cost, what happens when a learner’s explanation is wrong, and who checks it when there’s no teacher in the room. This entry covers what the research says, how to build why-prompts into a lesson, and what to look for in your platform.

What the research says

Two related techniques

Elaboration, broadly, means connecting new information to what you already know. Weinstein, Madan and Sumeracki, in their review of six well-supported learning strategies, define it in just those terms, and note that the word is used in several ways in the research (Weinstein et al., 2018). Two specific techniques sit under it.

Elaborative interrogation asks learners to generate an explanation for why a stated fact is true. Weinstein and colleagues trace the technique to work by Pressley and colleagues in 1987, and describe it as the technique under the umbrella of elaboration for which there is relatively strong evidence.

Self-explanation asks learners to explain to themselves how new information fits with what they know, or why each step of a solution was taken. Bisra and colleagues define it as “a process by which learners generate inferences about causal connections or conceptual relationships” (Bisra et al., 2018).

The difference is mostly in the material. Why-questions about stated facts suit factual content: a history course, a compliance module, an introduction to nutrition. Self-explanation prompts suit processes, worked problems and texts that build an argument. In a course, you’ll use both, and the design advice below applies to either.

How well supported is it?

The most cited assessment is the 2013 review by Dunlosky, Rawson, Marsh, Nathan and Willingham, which rated ten learning techniques on how widely their benefits held up across learners, materials and tests. In the authors’ words, “Elaborative interrogation, self-explanation, and interleaved practice received moderate utility assessments” (Dunlosky et al., 2013). Only practice testing and distributed practice were rated high. The review’s abstract also gives the main reason for the middle rating: both techniques “have not been adequately evaluated in educational contexts”.

The strongest quantitative evidence is for self-explanation. Bisra and colleagues pooled 69 effect sizes from 64 research reports, covering 5,917 participants, in which learners were prompted or directed to self-explain while studying or solving problems (Bisra et al., 2018). The overall effect was g = 0.55, a medium effect, with a 95% confidence interval of 0.45 to 0.65. They coded 20 possible moderators, including the type of task (solving problems, studying worked examples, studying text), subject area and level of education.

Two of their findings matter directly for course design. First, prompting learners to explain beat giving them an explanation: studies comparing self-explanation with instructional explanations found an advantage of g = 0.35 for self-explanation. A well-written explanation from you is not a substitute for the learner building their own. Second, the format of the prompt mattered. Multiple-choice prompts, where learners picked an explanation from a list rather than writing one, had an effect of g = 0.24 that was not statistically different from zero. Prompts that asked for conceptual explanations did better than prompts that asked learners to reflect on their own thinking.

A classic study, and its price

The study most often cited for self-explanation is by Michelene Chi and colleagues (Chi et al., 1994). Fourteen eighth graders read a 101-sentence passage on the circulatory system from a high-school biology text and were prompted to explain each line to themselves as they went. Ten others read the same text twice without prompts. Everyone was tested before and after.

Prompted to explain each line, students gained 32 points against 22 for reading twice

Gain from pretest to posttest in eighth graders' understanding of the circulatory system: 14 students prompted to self-explain while reading, 10 who read the same text twice · gain, pretest to posttest (% correct)

Prompted to self-explainRead the text twice

Both groups improved, but the prompted group improved more, and the advantage held on the harder questions that asked students to infer new knowledge or apply it to health problems. Within the prompted group, the students who generated the most explanations did best.

The detail the popular summaries leave out is time. The reading-twice condition was meant to roughly match the prompted group’s time on task, but it didn’t: the prompted students averaged about 2 hours 5 minutes with the text, against about 1 hour 6 minutes for the students who read it twice. Some of the benefit may simply be more time spent thinking about the material. That’s not a reason to dismiss the technique, because the time is spent doing something useful, but it’s a cost you pay in every lesson where you use it. Bisra and colleagues looked at this too: where the self-explanation group spent longer than the comparison group, the effect was larger than where the time was similar (g = 0.41), but the difference wasn’t statistically significant.

When explanations go wrong

The second thing the popular guides underplay is what happens when the learner’s explanation is wrong. Weinstein and colleagues are direct about it.

They also note that elaborative interrogation takes expertise to do well, takes a long time, and leaves open questions: whether learners need to generate their own why-questions or can answer ones provided for them, and whether the time is efficiently spent. Their conclusion is cautious: without answers to those questions, “it may be too early for us to instruct teachers to use this technique in their classes” (Weinstein et al., 2018).

How strong is the evidence?

Moderate. The self-explanation meta-analysis is large and the average effect is meaningful, but the effect varied with the kind of prompt, and the classic studies didn’t hold time on task equal. Dunlosky and colleagues judged that neither technique has been tested enough in real educational settings, and Weinstein and colleagues stop short of recommending elaborative interrogation for classrooms. We’d treat both as worth building in for the ideas that matter most, not on every screen, and always with a way for learners to check what they wrote.

How to apply it in a course

The question to ask of every key idea in a lesson is: would a learner who understood this be able to say why it’s true? If yes, ask them to say it, then show them a good answer.

We’ll use one example throughout: a module in an online food safety course for café staff, on storing and reheating cooked food. It’s an invented example, but it’s typical of compliance content, where learners can repeat a rule without understanding it, and then break it when the situation changes slightly.

1. Choose the facts worth explaining

Don’t put a why-prompt on every screen. Given the time cost, pick the two or three ideas per lesson that learners must understand rather than just recognise: the ones that transfer to new situations. In the food safety module, “cooked rice is a particular risk if left out” is worth a why. “The probe thermometer is kept in the second drawer” isn’t. Your learning objectives are the list to choose from: anything that asks learners to explain, apply or decide is a candidate.

2. Teach first, then ask

Elaboration links new information to what learners already know, so they need enough to link it to. Ask after the explanation or worked example, not before. A why-prompt on a topic learners have never met invites guessing, and guesses are where wrong explanations come from. For complete beginners, a short example first, then the prompt, works better than the prompt alone.

3. Write the prompt as a specific why or how

“Reflect on what you learned” is too vague to produce a useful explanation, and Bisra and colleagues found prompts asking for conceptual explanations did better than prompts about learners’ own thinking. Name the fact and ask for the reason: “Why is cooked rice riskier than bread left out for the same time?” For a process, ask about a step: “Why does the reheating rule say piping hot all the way through, not just hot?”

4. Ask learners to write, not pick

Multiple-choice prompts had the weakest effect in the meta-analysis, and the result wasn’t distinguishable from zero. Let learners type two or three sentences, or record a short voice note. A choice of explanations still has a place as a check after they’ve written their own, which connects this to writing multiple-choice questions that test reasoning rather than recall.

5. Show a model answer straight away

This is the step a self-paced course can’t skip. There’s no teacher to catch a wrong explanation, so after the learner submits, show a short model answer and ask them to compare: “Does your answer mention the spores surviving cooking? If not, read this again.” Write the model answer as carefully as the lesson itself: two or three sentences that give the reason, not a paragraph that repeats the rule. Our entry on effective feedback covers why an explanation beats a simple right-or-wrong.

6. Budget the time

If a why-prompt takes two to three minutes to answer properly, three per lesson adds most of ten minutes. Say so in the lesson’s estimated time, and keep the rest of the lesson lean so learners aren’t facing a long screen of content and a writing task at the end. That’s a cognitive load decision as much as a scheduling one.

How the platforms support it

Elaborative interrogation needs three things from a platform: somewhere for learners to type an answer, a way to show a model answer after they’ve submitted, and ideally a way for you to read what they wrote. We haven’t yet published our reviews of Thinkific, LearnWorlds, Teachable and Kajabi, and features change often, so check these in your platform’s current documentation, on the plan you’re paying for:

• Open-text questions inside a lesson. Can learners type a free-text answer in the middle of a lesson, rather than only in a quiz at the end of a module?
• A model answer after submitting. Can you show feedback text to every learner once they’ve answered an open question, even when it isn’t marked right or wrong?
• Reading the answers. Can you see and export what learners wrote? Their explanations are the best record you’ll get of where your teaching isn’t landing.
• Prompts inside video. Can you pause a video at a set point for a question, so the why comes straight after the explanation it’s about?

If your platform can’t do the second of these inside a lesson, a workable fallback is a short written prompt in the lesson text, followed by a “compare your answer” toggle or the model answer on the next screen.

Further reading

• Dunlosky et al. (2013), “Improving students’ learning with effective learning techniques”. The review that rated ten common study techniques, with long sections on elaborative interrogation and self-explanation and an honest account of what hasn’t been tested.
• Bisra et al. (2018), “Inducing self-explanation: A meta-analysis”. The best single source on how much self-explanation prompts help and which kinds of prompt work.
• Chi et al. (1994), “Eliciting self-explanations improves understanding”. The classic study, worth reading for its examples of what good and poor explanations look like.
• Weinstein, Madan & Sumeracki (2018), “Teaching the science of learning”. A clear review of six strategies, with the most useful cautions about elaboration. Open access.

For the rest of the principles behind good course design, browse our research library.