Elaborative interrogation for MRCPCH revision: how to ask why

Use elaborative interrogation selectively: choose one important, source-backed claim; hide the explanation; ask why, how, what-if or what-if-not; build the shortest defensible explanation from memory; then check every link against the source and test the idea on a fresh prompt. Asking "why?" is useful only when it exposes or strengthens a relationship. An unsupported story that merely sounds plausible is not learning evidence.

This article covers non-clinical MRCPCH examination preparation. It contains no clinical teaching, diagnostic, treatment, prescribing, dosing, safeguarding or patient-management guidance. Research on elaborative interrogation spans different ages, subjects, materials and outcomes; it does not establish that this technique improves MRCPCH scores or pass rates.

The direct answer: ask a question that can fail

Elaborative interrogation is a learning strategy in which you generate an explanation for why a stated fact or relationship is true. Broader deep-reasoning prompts can also ask how something works, what would change if one condition changed, or why a tempting alternative does not follow.

The word why is not the active ingredient by itself. A useful prompt forces you to connect the claim with relevant prior knowledge, state a mechanism or relationship and reveal where the explanation stops. A weak prompt invites a slogan, an unsupported association or an answer copied from the line above it.

A good elaborative question has a checkable target, requires a relationship rather than a definition, and leaves you able to distinguish a source-backed explanation from a plausible invention.

Use the DEEP loop: Define the claim, Explain from memory, Evaluate against the source and Prove it elsewhere.

Key points

• Reviews and meta-analyses report positive average effects for elaborative interrogation, but results vary substantially across learners, domains, materials and tests. The method is promising, not a guaranteed shortcut or an MRCPCH outcome predictor.
• Prior knowledge matters. If you cannot produce a credible explanation, first build a small accurate foundation; otherwise a "why" answer may become confident speculation.
• The explanation needs feedback. Check the claim, causal links, conditions and exceptions against an authoritative source, then use delayed retrieval or a fresh question to test whether the relationship is usable.

What the evidence actually supports

Early experiments showed a fact-memory benefit

Pressley and colleagues asked university students to generate reasons why unfamiliar facts made sense. Across four experiments, elaborative interrogation and imagery produced better memory than a reading-to-understand control. These studies helped establish the method, but their materials and immediate learning conditions should not be treated as a direct model of an MRCPCH examination.

The practical lesson is narrower: searching for a meaningful relation can make a fact less isolated. It does not follow that adding "why?" to every line of a textbook will improve every kind of learning.

Prior knowledge changes the result

Woloshyn, Pressley and Schneider compared elaborative interrogation with reading across familiar and unfamiliar domains. Prior knowledge strongly affected learning. Elaborative interrogation helped across conditions, but learners without relevant knowledge did not reach the performance of knowledgeable learners who simply read the material.

A separate study by Willoughby and colleagues found an advantage over repetition for familiar animals on immediate and one-month tests, but not for unfamiliar animals. Together, these findings support a simple boundary: elaboration needs something accurate to elaborate with.

If your answer is "because that is the rule", or if each link is guessed, pause. Read a concise authoritative explanation, identify the relationship, then retry the question with the source closed.

Conceptual and inferential learning can benefit, but not invariably

McDaniel and Donnelly studied short descriptions of scientific concepts. Elaborative interrogation improved both factual and inference performance in their experiments, including when compared with labelled schematics. This is useful evidence that the method can reach beyond isolated fact matching.

However, O'Reilly and colleagues later compared elaborative interrogation, self-explanation and repetition for scientific facts. In that study, self-explanation outperformed the other conditions, while elaborative interrogation did not outperform repetition. A learning technique should therefore be selected for the actual knowledge problem, not defended as universally superior.

Reviews describe moderate support with important uncertainty

Dunlosky and colleagues rated elaborative interrogation as a moderate-utility technique: evidence was encouraging, implementation was simple and effects appeared across several conditions, but important gaps remained around generality and long-term retention. Donoghue and Hattie's later meta-analysis reported a positive average effect with substantial heterogeneity. That heterogeneity matters more than a league-table rank: different prompts, materials, learners and outcomes can produce different results.

The US National Academies' How People Learn II places elaborative interrogation among explanation-generating strategies and stresses deep questions such as why, how, what-if and what-if-not. It also notes that learners often need training to ask productive questions and that much experimental work used isolated facts or relatively short texts.

Distinguish elaborative interrogation from nearby methods

Retrieval practice asks "Can I produce it?"

Retrieval practice requires an attempt to recall information or use it without the answer displayed. Elaborative interrogation can include retrieval, but adds a relational demand: why does the claim hold, how do the parts connect, or what would change under another condition?

If the immediate problem is simply that a label, list or rule cannot be recalled, a direct retrieval prompt may be more efficient. The closed-book revision guide explains how to separate learning input from a genuine memory check.

Self-explanation asks "What does my reasoning mean?"

Self-explanation is often broader and may cover the steps used to solve a problem, how new information fits prior knowledge or why a chosen step follows. Elaborative interrogation usually begins with a more specific question about a stated claim or relation.

The difference is not a rigid border. Use the prompt that exposes the missing link. Do not spend time classifying a useful explanation merely to give the activity a perfect label.

The Feynman technique asks for an audience-ready explanation

The Feynman technique guide uses a bounded teaching task: explain an idea plainly, expose gaps, correct them and test transfer. Elaborative interrogation is smaller-grained. One "why" question may repair one relation inside that larger explanation.

Summarising asks "What is essential?"

The summary-notes guide focuses on selection and compression. An elaborative question can strengthen one claim in a summary, but a paragraph of causal reasoning is not automatically a good summary, and a concise summary is not automatically an explanation.

Choose a question with a real learning job

Ask why when a claim feels arbitrary

Use "Why can this be true?" when two facts are repeatedly confused or when a rule is remembered only as a detached sentence. The aim is to uncover the relationship that makes the claim less arbitrary.

Do not force a causal explanation onto a convention, label or externally set rule. For those, ask what the rule is, where the current authoritative version lives and how you will retrieve it.

Ask how when a sequence or process is missing

"How does A lead to B?" requires intermediate steps. Score the answer by whether every step is necessary, accurate and ordered - not by its length.

If the source does not establish the mechanism, say so. Filling a gap with a smooth story is worse than marking the relationship as unresolved.

Ask what-if when transfer is the goal

"What would change if this condition were absent?" or "What would remain true?" tests whether you understand the boundary of a relationship. Keep this within material supported by the source; do not manufacture clinical scenarios or recommendations.

Ask what-if-not to test an alternative

"Why does the tempting alternative not follow?" can expose the exact distinction being missed. The answer should identify a discriminating condition, not merely repeat that the alternative is wrong.

Ask compared-with-what when a claim is too broad

Statements such as "this method is effective" are incomplete without a comparator, outcome and interval. Ask: effective compared with what activity, for which learners, on what test and after what delay? This is particularly useful when reading learning-science headlines.

Use the DEEP loop

D - Define one claim and one purpose

Write one sentence that can be checked. Then state the reason for interrogating it: repeated confusion, missing mechanism, weak transfer or an over-broad claim.

Poor target: "Why is spaced revision good?"

Better target: "Why might distributing the same study time across separated sessions improve delayed retrieval compared with one massed session?"

The improved version identifies a comparator and an outcome. It still needs a source-backed answer; it is not a licence to invent a cognitive mechanism.

E - Explain from memory before reopening the source

Hide the explanation and write or say the shortest chain you can defend. Mark uncertainty explicitly. A useful answer may be three sentences; a page of prose can conceal the missing link.

If nothing credible appears within a minute or two, that is diagnostic information. Build a small foundation rather than generating increasingly elaborate guesses.

E - Evaluate every link

Reopen the authoritative source. Check:

1. Is the starting claim represented accurately?
2. Does the source support each link in the explanation?
3. Have association and causation been confused?
4. Are conditions, populations, comparisons or dates missing?
5. Did the explanation introduce a new claim that needs its own source?

Correct in a visibly different colour or version so that the gap remains identifiable. Do not replace the attempt with a clean copied paragraph; that removes the evidence needed to plan the next return.

P - Prove the relationship elsewhere

Close the source again. Use one of four checks:

• answer the same "why" prompt after a delay;
• explain the relation in the reverse direction where that is logically valid;
• answer a changed what-if question;
• apply the relation to a genuinely first-seen, independently authored question.

A familiar explanation may become fluent without becoming transferable. The proof step asks whether the relation can be retrieved and used beyond the original wording.

A worked non-clinical example

Suppose a candidate notices that repeated question-bank items make the displayed percentage rise. The claim to examine is:

A higher mixed score may not represent stronger readiness when familiar repeats and first-seen items are combined.

The initial "why" explanation might be: familiar questions contain memory cues from previous attempts, so the score partly measures item familiarity rather than performance on unseen decisions. When the source is checked, the candidate should verify what the platform actually labels, avoid claiming that every repeat is worthless and preserve the boundary: repeated questions can support retrieval, but their scores answer a different question from first-seen performance.

A what-if test then asks: "What would the dashboard show if first-seen and repeated items were reported separately?" The candidate can use the question-bank score guide to distinguish performance signals without turning any percentage into a pass prediction.

This example works because the prompt has a checkable relationship and a practical consequence. It does not require clinical content or recalled examination material.

Common failure modes

Asking why about everything

Interrogating every sentence creates a large review burden and treats trivial, definitional and relational knowledge alike. Reserve the method for important claims that feel arbitrary, unstable or difficult to transfer.

Accepting the first plausible story

Fluency and coherence are not accuracy. Require a source check and mark unsupported links. If the authoritative source states the association but not a mechanism, do not upgrade it to causation.

Starting before there is enough knowledge

Prior-knowledge studies suggest that elaborative interrogation is less dependable when the learner lacks a relevant knowledge base. Use a short explanation, worked example or pretesting step to diagnose the gap, then retry.

Writing an essay instead of testing a relation

Long answers increase opportunity for unsupported detail. Set a sentence or time limit, then score the explanation by accuracy, necessary links and boundaries.

Never returning to the question

The first explanation is an encoding event, not proof of durable learning. Schedule one delayed prompt. The blurting guide offers a broader closed-book check when several connected relations need review.

Using the method when tired enough to guess

Elaborative interrogation demands effortful generation and evaluation. A 2019 experiment specifically examined resource depletion and found that learner state can alter the costs and benefits. If explanations become vague or source checking becomes superficial, switch to a bounded lower-demand task or stop. Productive difficulty is not unlimited difficulty.

Use elaborative questions with PaedVault

After an independent question attempt and feedback review, select at most one relation worth interrogating. Write the prompt in your own words without copying question text, options or explanations. Answer from memory, check the underlying principle against an authoritative source and test it later on a different independently authored item.

Do not upload PaedVault content to another service or share protected material with a study group. The PaedVault Acceptable Use Policy limits content to personal use. A useful personal prompt contains the learning relationship, not a reconstruction of the protected question.

Pair elaborative interrogation with existing tools rather than creating another large note system:

• add one checked "why" prompt to a small flashcard only when a relational answer is scorable;
• attach a question to a diagram when the relation is genuinely spatial, sequential or comparative;
• use it during interleaved revision to distinguish two methods or structures that are repeatedly confused;
• retire the prompt when delayed retrieval and first-seen application are stable.

A seven-day comparison

Test the method on your own workflow without treating the result as a controlled trial.

1. Select six non-clinical learning claims or exam-logistics relationships that are important and currently unstable.
2. Match them by difficulty as closely as practical.
3. For three, use DEEP: define, explain, evaluate and schedule a proof.
4. For three, use your normal source review for the same total time.
5. After two days, answer a short retrieval prompt without notes.
6. After seven days, use a changed why, how or what-if question.
7. Record accuracy, unsupported links, time spent and whether the relation transferred to a fresh task.

Keep the technique only where it adds information or transfer beyond simpler retrieval. Preference and effort are observations, not outcomes.

Conclusion

Elaborative interrogation can make an isolated claim more meaningful by requiring a learner to generate and check the relationship behind it. The evidence is encouraging but conditional: prior knowledge, prompt quality, source feedback, material and outcome all matter, and no research establishes an MRCPCH pass-rate benefit.

Use DEEP: Define one claim, Explain from memory, Evaluate every link and Prove the relation elsewhere. Ask fewer, better questions. A source-backed explanation that survives a changed prompt is useful; a fluent "why" story that cannot be checked is not.

Frequently asked questions

What is elaborative interrogation in MRCPCH revision?

It is the deliberate use of a deep question - usually why, how, what-if or what-if-not - to generate an explanation for a source-backed claim or relationship. For MRCPCH revision, keep the prompt non-clinical unless qualified review is available, answer from memory, check every link against the authoritative source and test the relationship again later.

Does asking why improve memory?

Studies and reviews report positive average effects in several settings, especially for making facts meaningful, but results vary and are not MRCPCH-specific. Benefits depend on the material, prior knowledge, prompt and assessment. Asking "why?" alone is insufficient: the explanation must be accurate, checked and later retrieved or applied.

How many why questions should I write?

There is no evidence-based MRCPCH quota. Start with one question for an important, recurring or poorly understood relationship. If generating and checking explanations displaces first-seen questions, feedback, sleep or syllabus coverage, reduce the number. Quality and later usability matter more than a growing prompt list.

What if I cannot answer the why question?

Treat that as a diagnostic result. Read a concise authoritative explanation, identify the missing relation and retry with the source closed. If the source does not establish the mechanism, record the narrower supported claim rather than inventing an answer. Prior knowledge is an important condition for useful elaboration.

Is elaborative interrogation the same as the Feynman technique?

No, although they overlap. Elaborative interrogation usually targets one stated relationship with a specific why, how or counterfactual prompt. The Feynman technique asks for a broader audience-ready explanation of a bounded idea. Both require retrieval, source checking and correction; choose the smaller task that exposes the actual gap.

Sources

1. Royal College of Paediatrics and Child Health. Theory exams: structure and syllabi. Updated 27 August 2026.
2. Dunlosky J, Rawson KA, Marsh EJ, Nathan MJ, Willingham DT. Improving students' learning with effective learning techniques. Psychological Science in the Public Interest. 2013;14(1):4 - 58. doi:10.1177/1529100612453266.
3. Donoghue GM, Hattie JAC. A meta-analysis of ten learning techniques. Frontiers in Education. 2021;6:581216.
4. National Academies of Sciences, Engineering, and Medicine. How People Learn II: Learners, Contexts, and Cultures - Knowledge and Reasoning. Washington, DC: The National Academies Press; 2018.
5. Pressley M, Symons S, McDaniel MA, Snyder BL, Turnure JE. Elaborative interrogation facilitates acquisition of confusing facts. Journal of Educational Psychology. 1988;80(3):268 - 278.
6. Woloshyn VE, Pressley M, Schneider W. Elaborative-interrogation and prior-knowledge effects on learning of facts. Journal of Educational Psychology. 1992;84(1):115 - 124.
7. Willoughby T, Waller TG, Wood E, MacKinnon GE. The effect of prior knowledge on an immediate and delayed associative learning task following elaborative interrogation. Contemporary Educational Psychology. 1993;18(1):36 - 46.
8. McDaniel MA, Donnelly CM. Learning with analogy and elaborative interrogation. Journal of Educational Psychology. 1996;88(3):508 - 519.
9. O'Reilly T, Symons S, MacLatchy-Gaudet H. A comparison of self-explanation and elaborative interrogation. Contemporary Educational Psychology. 1998;23(4):434 - 445. doi:10.1006/ceps.1997.0977.
10. Kühl T, Bertrams A. Is learning with elaborative interrogation less desirable when learners are depleted?. Frontiers in Psychology. 2019;10:707. doi:10.3389/fpsyg.2019.00707.
11. PaedVault. Acceptable Use Policy. Updated 28 July 2026.