FRCR 2A Pass Mark: How Standard Setting Works

The RCR does not publish one permanent pass percentage for every FRCR Part 2A sitting. The required standard is determined for the examination being sat, so a number quoted from a previous cohort should not be treated as the pass mark for a future sitting.

This is an important distinction. A criterion-referenced examination aims to identify whether candidates meet the expected standard of a just-passing candidate; it is not designed simply to pass a fixed proportion of the cohort or to use one unchanging raw percentage forever.

The RCR states that it uses a modified Angoff method. A panel of subject-matter experts considers each question and estimates how likely a just-passing candidate would be to answer it correctly. Those judgements are reviewed and discussed before the standard for the paper is calculated.

In practical terms, an easier set of questions should require a higher raw score than a harder set of questions to represent the same underlying standard. That is why the difficulty of an individual paper matters and why one historical percentage cannot be applied mechanically to every sitting.

The final pass mark is determined after the examination using the Hofstee compromise method. The RCR explains that this takes account of candidate performance alongside the judgements made by the standard setters. Candidate cohort and pass rate can therefore vary between sittings.

This does not mean the mark is chosen arbitrarily after seeing the results. It means that expert judgement and observed examination performance are brought together within the published standard-setting process.

Because there is no penalty for an incorrect answer, leaving a question unanswered cannot improve your raw mark. Candidates should therefore plan to offer an answer to every item, even when they have to choose between two plausible options.

After each sitting, expert examiners conduct statistical review and quality-control checks. The RCR states that poorly performing questions are scrutinised and may be removed where error or material ambiguity is identified. If questions are removed, the pass mark is adjusted so candidates are not disadvantaged.

A difficult question is not automatically a defective question. The relevant issue is whether the item is fair, correctly keyed and functioning as an appropriate assessment of the published curriculum.

There is no honest universal percentage that can be promised as a guaranteed pass. Question banks differ in difficulty, candidate populations differ and repeated exposure can inflate scores. Instead of chasing one magic number, look for evidence that your performance is becoming broad, stable and less dependent on familiar questions.

A useful mock result separates answered, incorrect and unanswered questions. A low raw score caused by many blanks is different from a low score after completing all 120 questions. Likewise, a correct answer reached by a fragile guess should not be treated as secure knowledge simply because it added one mark.

Review the paper by error type: knowledge gap, missed imaging discriminator, incorrect task interpretation, timing problem or overconfident guess. Then convert the pattern into a specific revision action rather than repeating the whole curriculum indiscriminately.