Spearman correlation — MRCPsych Paper A MCQ

Instant feedback + full explanation. One question, done properly.

EasyEpidemiology and research methodsSpearman correlationMRCPsych Paper A

A researcher records an ordinal side-effect severity score and an ordinal medication adherence score for each participant. The data are non-normally distributed. Which statistical test is most appropriate for assessing whether there is a monotonic association between the two variables?

Educational content. Not a substitute for clinical judgement or local policy.

Reveal the answer and explanation

Correct answer: D — Spearman rank correlation

Spearman rank correlation is the best answer because it assesses the strength and direction of a monotonic association using ranked observations. It is suitable when variables are ordinal or when continuous data do not satisfy the assumptions required for Pearson correlation. Pearson correlation primarily assesses linear association between continuous variables and is less appropriate here. An unpaired t-test compares the means of two independent groups rather than association between paired variables. A chi-squared goodness-of-fit test compares observed categorical frequencies with an expected distribution. Cox regression models time-to-event outcomes and requires survival or censoring data, neither of which is present.

Reference: The BMJ, Statistics at Square One, section 11: Correlation and regression—Spearman rank correlation. https://www.bmj.com/about-bmj/resources-readers/publications/statistics-square-one/11-correlation-and-regression