The first Canadian farm-protest organization was formed in Manitoba. In an early-1990s memory study, learners asked “Why does that make sense given what you know about that particular province?” learned facts like this one better. Learners asked why such a fact was unexpected, given what they knew about other provinces, showed no gain (Martin & Pressley, 1991, as cited in Woloshyn et al., 1992).
Both are why questions and both take effort, but they send you in opposite directions: the first toward what you already know about Manitoba, the second away from it. That contrast is elaborative interrogation in miniature. The strategy means pausing at a stated fact and generating an explanation for why it is true. The post on elaboration treated connecting new material to old knowledge as a broad habit. This is one move inside that family, a generative strategy, meaning you produce the explanation yourself instead of reading one. Because the move is so specific, researchers have been able to test where it pays off and where it mostly costs time.
The Animal Study That Isolated the Ingredient
Willoughby, Waller, Wood, and MacKinnon (1993) designed a study that could tell apart two explanations: that asking why helps because it takes effort, or because it connects to what you already know. Introductory psychology students learned 60 short facts about ten animals. Five animals were familiar, such as the house mouse, and five were unfamiliar, such as the coati. One group answered aloud, after each fact, “Why would that animal do/have that?” A second group read each fact aloud, with instructions to understand it. After a short unrelated task, everyone matched facts to animals, and did so again four weeks later.
Asking why beat reading for the familiar animals only, and the advantage held a month later. For unfamiliar animals, the groups did equally well.
If the benefit came from effort or general alertness (arousal), it should have appeared for any animal. If anything, the authors argued, effort might have been expected to be higher for the unfamiliar ones. Yet the gain appeared only where students had knowledge to draw on. Their conclusion was that “it is not active learning per se that is responsible for elaborative interrogation benefits, but the making of connections to prior knowledge” (Willoughby et al., 1993, p. 44).
A year earlier, a second study had found something less tidy. Woloshyn, Pressley, and Schneider (1992) had Canadian and West German undergraduates learn facts about Canadian provinces and West German states, so every student who studied met one familiar set and one foreign set. Asking why beat reading in all four combinations of student and country, foreign facts included. The advantage looked larger for home-country facts, but the study’s main statistical tests did not confirm that difference. Prior knowledge was the strongest single predictor of scores. Students asking why about the foreign country still scored lower than home-country students who had only read. For the foreign-fact gain, the authors pointed to “general factors such as increased arousal, attention, and cognitive effort” (p. 121).
Read side by side, the two studies point to a gradient rather than an on-off switch. Dunlosky et al. (2013, p. 9), reviewing the wider literature, describe effects that “generally increase as prior knowledge increases” and call the benefit for lower-knowledge learners “less certain.”
When the “Why” Goes Wrong
Willoughby’s team also sorted every answer students gave. Unfamiliar animals more often drew answers that were wrong or what the authors called “pat,” meaning too general to single out one animal. One such answer explained that a skunk eats corn because corn is available where it lives, which is true of almost any animal that eats corn. Facts given pat answers were matched less often than facts given correct, specific ones, and pat answers “did not help the students to distinguish among the animals” (p. 45).
Yet in the same study, matching was no better after adequate answers than after inadequate ones, or after no answer at all. In Woloshyn’s study, foreign facts almost never drew an adequate answer, and asking why still beat reading. Dunlosky et al. (2013, p. 9) call the role of quality “still an open question,” partly because these analyses sort items after the fact by how learners happened to answer.
Sometimes the strategy loses to plain rereading. Clinton, Alibali, and Nathan (2016) taught posterior probability, the reasoning behind what a positive medical test actually means, to undergraduates who could not yet solve such a problem. Students read a lesson of about a thousand words. Some read it twice, some answered simple embedded questions, and some answered harder how and why questions whose answers had to be pieced together from the lesson. In the same session, everyone then solved four problems of the same kind.
The how-and-why group spent longer and rated the lesson harder, yet solved fewer problems than the students who read twice. Nearly nine in ten of their answers were missing, circular, wrong or inadequate. Students who wrote better answers scored higher, though a correlation like that cannot show that the answers caused the scores. The authors offered two explanations, poor answers and difficulty that may have “overwhelmed” students, and judged it “premature to propose implications” (Clinton et al., 2016, p. 593).
Kühl and Bertrams (2019) found no benefit with different material. Ninety-six participants, nearly all university students, read a text of about 300 words on how airplanes produce lift. For half of them, explanatory segments were removed and replaced by prompts asking them to generate the missing information. These learners answered roughly 80% of the prompts correctly. Even so, they did no better than readers of the full text on a retention test given almost straight away, or on transfer questions that applied the ideas to new problems. They spent about three times as long studying and reported more mental effort. The authors left open whether the strategy needs longer materials to show its value.
Both studies expected the strategy to act as a desirable difficulty, the kind of effort that feels costly during study and pays off later (the subject of an earlier post in this series). Neither found the later payoff.
What the Evidence Covers, and What It Doesn’t
The positive evidence is about memory for the facts you questioned. Dunlosky et al. (2013) rate the strategy’s overall usefulness as moderate, with the firmest support for lists of discrete facts, like the animals and provinces above. Effects have also turned up for facts inside longer texts. In one classroom study, biology lab undergraduates studied a textbook passage on digestion, half of them with 21 why prompts spread through it. The prompted group did better on a true-or-false test, and the prompts added about four minutes to a half-hour session (B. L. Smith et al., 2010, as cited in Dunlosky et al., 2013). The same review warns that effects can be diluted or even reversed when prompts are sparse in a long text. Add the Clinton and Kühl results, and the record for complex, connected material is mixed.
The record is thinnest on transfer. Remembering which fact goes with which animal is retention. Using an idea on a problem you have not seen before is transfer, and for this strategy it has hardly been tested. The nearest evidence is three studies in Dunlosky’s review that went beyond memory for the stated facts, using questions that required inference or integration. All three found benefits on those questions. One of the three also gave a problem-solving test and found no benefit there. Kühl and Bertrams found none on their transfer questions either. Any claim that asking why will help you apply what you learn runs ahead of that record.
The record is also short. Most studies tested memory within minutes. Where benefits were reported weeks or months later, including Willoughby’s one-month result, an earlier test usually came first, and that test may itself have acted as retrieval practice (Dunlosky et al., 2013). Elaborative interrogation is something you do while studying. Whatever head start it gives a fact, the earlier posts on retrieval and spacing cover what keeps that fact usable.
Using It Well
Most studies use some version of one prompt: “Why would this fact be true of this [X] and not some other [X]?” (Dunlosky et al., 2013, p. 8). The second half of that question does real work. On the review’s account, linking a fact to what you know is only part of the job: learners “must also be able to discriminate among related facts” (p. 8). To take an untested illustration from pharmacology, a medical student might ask why one drug causes a particular side effect when the drug next to it in the same class does not.
Aim the strategy at discrete facts in a subject where you already have a foothold. A quick check for a pat answer: would it explain the same fact just as well for the neighboring animal, drug or disease? If so, take that as a sign you need more of the foundation first. On hard conceptual material you are meeting for the first time, Clinton’s probability study favored simply reading the lesson twice. Check every answer against your text, your notes or a teacher. Weinstein et al. (2018) give the same advice, warning, on the strength of the Clinton study, that poor elaborations can hurt learning. And watch the clock: in the airplane study, the prompts roughly tripled study time and bought nothing measurable.
For people who teach, the same studies carry a caution. Clinton et al. (2016) suggested that teachers model good answers and monitor their quality. Weinstein et al. (2018) list open questions about how much expertise students need and how long they should search for an answer, and conclude that “it may be too early for us to instruct teachers to use this technique in their classes” (p. 11). They also ask whether a class might do better to answer a few questions, pool what they find and then move on to retrieval practice.
What This Looked Like for Me
Asking why came naturally to me, as I described in the elaboration post. These studies helped me see why the habit paid off in some places and not in others.
Medical training has an obvious version of the animal study. When a new fact sits on top of something you already understand, such as a disease mechanism built on physiology you have already learned, a why question has something to grab. When the fact arrives before its foundation, as in the first days of an unfamiliar subject, the same question tends to produce a pat answer that is plausible, general and attached to nothing. That kind of studying feels productive and, on this evidence, adds little.
The lesson I take from this is sequencing: a first read and some basic retrieval to build the base, then the why questions, with answers checked against the text, the next round of practice questions or an answer key.
I did not run a controlled experiment on myself. I used this habit alongside retrieval practice, spacing and other strategies, and I cannot isolate what it contributed to my board scores.
The Core Lesson
Asking why a fact is true can strengthen your memory for it, most reliably when you already know a good deal about the topic. With little prior knowledge the benefit becomes uncertain, and in one study of hard conceptual material it did worse than reading twice. Its record is on retention soon after study, and the evidence on transfer and long delays is thin. Use it once you have a foothold, and treat a vague answer as a sign to go back to the material.
The next post turns to a close relative, self-explanation, which asks why each step follows in a worked example or procedure.
This post is part of a series on the science of learning. Each post covers one evidence-based principle and how to apply it to your own studying. Follow us on Instagram @acemeded to keep up with future blog posts and related content.
References
Clinton, V., Alibali, M. W., & Nathan, M. J. (2016). Learning about posterior probability: Do diagrams and elaborative interrogation help? The Journal of Experimental Education, 84(3), 579–599. https://doi.org/10.1080/00220973.2015.1048847
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. https://doi.org/10.1177/1529100612453266
Kühl, T., & Bertrams, A. (2019). Is learning with elaborative interrogation less desirable when learners are depleted? Frontiers in Psychology, 10, Article 707. https://doi.org/10.3389/fpsyg.2019.00707
Weinstein, Y., Madan, C. R., & Sumeracki, M. A. (2018). Teaching the science of learning. Cognitive Research: Principles and Implications, 3, Article 2. https://doi.org/10.1186/s41235-017-0087-y
Willoughby, T., Waller, T. G., Wood, E., & MacKinnon, G. E. (1993). The effect of prior knowledge on an immediate and delayed associative learning task following elaborative interrogation. Contemporary Educational Psychology, 18(1), 36–46. https://doi.org/10.1006/ceps.1993.1005
Woloshyn, V. E., Pressley, M., & Schneider, W. (1992). Elaborative-interrogation and prior-knowledge effects on learning of facts. Journal of Educational Psychology, 84(1), 115–124. https://doi.org/10.1037/0022-0663.84.1.115
