Should you change your first MRCPCH answer?

Change an MRCPCH answer when you can identify new, specific evidence that makes another option better - not simply because the first choice feels uncomfortable. Research across multiple-choice examinations generally contradicts a blanket "never change your first answer" rule, but it does not justify indiscriminate switching. Re-read the task, compare the final options and make one reasoned decision.

The useful rule is neither always trust your instinct nor always reconsider. It is: revise an answer when the reason for changing is stronger than the feeling that triggered the doubt.

Why the first answer feels safer than it is

Changing a correct answer to an incorrect one is memorable. It creates a sharp counterfactual: "If only I had left it alone." Keeping an incorrect answer when a later thought could have corrected it often feels less dramatic, even though the lost mark is the same.

This asymmetry helps explain the first-instinct fallacy. In a series of studies, Kruger and colleagues found that people tended to believe answer changing was harmful even though most observed changes moved from incorrect to correct. They argued that the regret attached to a damaging switch makes those cases easier to remember.

That finding does not mean your second thought is automatically superior. It means the emotional vividness of a wrong change is not reliable evidence for a universal first-answer rule.

For MRCPCH, the distinction matters because RCPCH currently uses single best answer questions. Five options may contain plausible elements, but one is superior for the task being asked. The best revision decision therefore needs a comparison, not an intuition about which answer came first.

Use the PaedVault MRCPCH exam guide for the current examination overview and follow its links to official RCPCH information near your sitting.

What answer-changing research can - and cannot - tell you

The broad direction of the evidence is reasonably consistent, but its limits are important.

A 2024 scoping review synthesised ten studies of answer changing in undergraduate and graduate objective examinations, with a focus on nursing education. It found that wrong-to-right changes were more common and concluded that changing answers did not generally reduce scores. The authors also highlighted variation in study methods, which limits simple predictions for an individual candidate.

A small 2005 study examined 36 answer books from a German high-stakes medical board examination containing five-option questions. Scores improved when candidates changed an answer once after reflection; additional changes did not improve scores. A later study of medical students also reported a net benefit from answer changes.

These are not MRCPCH trials, and paper marks, computer actions and candidate populations are not interchangeable. The evidence challenges "never change", but it does not establish an MRCPCH-specific score gain or a target number of changes.

The most defensible conclusion is narrower:

The order in which two answers occurred is not enough to choose between them. Change only when re-reading produces a reason that favours the alternative.

Use the CHANGE gate before revising an answer

When two options compete, run one short decision gate. CHANGE keeps the review anchored to the item rather than to rising anxiety.

C - Clarify the exact request

Read the lead-in again. What kind of answer is required? Do not solve a neighbouring question. A technically true option can still be inferior if it answers a different task.

H - Highlight the new cue

Name what you noticed after the first choice. Useful new evidence includes a previously missed word, a qualifier, a contradiction between the selected option and the stem, or recognition that the first choice answered the wrong question.

"This answer suddenly feels wrong" is not a cue. "I missed the word most" is.

A - Arrange a head-to-head comparison

Compare the original answer only with the strongest rival. Do not reopen all five options unless the original interpretation of the question has changed.

Complete two sentences:

• "The original option fits because…"
• "The alternative is better because…"

The second sentence must explain superiority, not merely plausibility.

N - Name the error in the first decision

If you change, identify what was wrong with the first process:

• task error: you answered a different lead-in;
• reading error: you missed or reversed a qualifier;
• knowledge retrieval: a relevant distinction became available;
• comparison error: you recognised why a plausible option was not the best; or
• marking error: the selected option did not match your intended answer.

If no error can be named, the change may be driven by discomfort rather than evidence.

G - Give one decisive reason

State the shortest reason that separates the two options. If the reason depends on adding facts not present in the stem, treat it cautiously. Single best answer decisions should be made from the presented task and evidence, not an imagined alternative scenario.

E - End the loop

Make the change or retain the original, then move on. Repeatedly alternating between the same options consumes time without necessarily adding evidence. A question can remain uncertain after a careful decision; uncertainty is not proof that another pass will resolve it.

Change for evidence, not for these three feelings

Three common sensations can trigger unproductive switching.

"I have chosen the same letter too often"

The pattern of preceding answer positions does not determine the next answer. Judge the current item. Attempts to manufacture a balanced sequence replace item evidence with an invented distribution rule.

"The obvious answer must be a trap"

A question may be difficult because several options are plausible, but difficulty does not prove deception. Ask which option best meets the lead-in. Do not reject a well-supported answer merely because it was recognised quickly.

"I still feel uncertain"

Confidence and accuracy are related imperfectly. Experimental work on answer revision suggests that the cause of an error matters: revision may correct a rushed response, but it may not resolve confusion between genuinely similar alternatives. Convert uncertainty into a named comparison. If no new discriminating reason emerges, retain the completed decision and continue.

The build-contrast-mix method can strengthen this skill during revision by making you state the feature that separates close alternatives.

Build a personal answer-change record in PaedVault

Population-level evidence cannot tell you whether your own changes are well targeted. Practice data can.

For several timed PaedVault blocks, mark every item considered for revision and record four fields before reading the explanation:

Field	What to record
Initial choice	The first selected option
Final choice	Whether you retained or changed it
Trigger	New cue, task correction, comparison, memory, or unevidenced doubt
Confidence	High, medium or low after the final decision

After marking, classify the outcome:

• wrong to right;
• right to wrong;
• wrong to wrong; or
• right retained / wrong retained.

Do not judge the strategy from one painful right-to-wrong switch. Review a series of blocks and ask whether changes made through the CHANGE gate differ from changes made through vague doubt.

The more useful metric is not "How often did I change?" It is "How often did I identify a valid reason before changing?"

Send task, reading and comparison errors to the incorrect-answer review workflow. Send unstable distinctions to the spaced-repetition queue so they are retrieved again after a delay.

Practise review as a separate exam skill

Answer revision should be rehearsed before exam day. In a timed question block:

1. answer each item using your normal process;
2. mark only items with a specific unresolved comparison;
3. return if the practice format and remaining time allow;
4. apply CHANGE once;
5. record the reason before viewing feedback; and
6. audit the outcomes after the block.

This separates two skills: solving the item initially and deciding whether later evidence justifies revision. A candidate can be knowledgeable yet change poorly because every low-confidence answer is reopened without a decision rule.

Use mock sessions to test how much review is realistic within the current paper format. Verify format and platform instructions on the official RCPCH pages rather than assuming that a practice interface exactly matches your sitting.

Do not confuse no negative marking with random changing

RCPCH currently states that all theory questions are worth one mark and that there is no negative marking. That supports answering every item: a wrong answer does not deduct an additional mark.

It does not imply that switching between answered options carries no opportunity cost. Time spent reopening one uncertain item cannot be spent checking a misread lead-in elsewhere. The practical distinction is:

• unanswered item: select the best available option because there is no negative marking;
• answered item with new evidence: compare and change if the alternative is superior;
• answered item with only renewed anxiety: retain the completed decision and move on.

This is an exam-process rule, not a claim about the clinical content of any question.

Conclusion

Do not protect an MRCPCH answer merely because it came first, and do not replace it merely because doubt appeared later. Re-read the request, identify the new cue, compare the strongest options and name the error in the first decision. Then make one reasoned choice and end the loop.

The first answer has no special status. The better-supported answer does.

Frequently asked questions

Is it bad to change answers in MRCPCH?

Not automatically. Research from other multiple-choice and medical examinations generally finds more wrong-to-right than right-to-wrong changes, but it is not MRCPCH-specific. Change when you can state new evidence that makes another option superior, not simply because confidence has fallen.

Should I always trust my first instinct in an SBA?

No. A blanket first-instinct rule is not supported by the broader evidence. Keep the first answer when re-reading produces no better reason; change it when you identify a task, reading, knowledge, comparison or marking error.

When should I review a flagged MRCPCH question?

Review a flagged item when it contains a specific unresolved comparison and the current examination instructions and remaining time permit. Practise this during timed blocks, and verify the live RCPCH platform guidance near your sitting rather than relying on an assumed review workflow.

Does MRCPCH use negative marking?

RCPCH currently states that its theory examinations do not use negative marking and that each question is worth one mark. Check the official structure page near your sitting in case the rules change.

Sources

• RCPCH theory exam structure and syllabi
• Changing test answers: a scoping review
• Answering multiple-choice questions in high-stakes medical examinations
• Answer changing in multiple choice assessment: change that answer when in doubt
• Counterfactual thinking and the first instinct fallacy
• Not all errors are created equal: metacognition and changing answers