Evidence-appraisal glossary
Simpson's paradox
Simpson's paradox is when a trend that appears in every subgroup of the data reverses or vanishes once the subgroups are combined. It typically happens because a lurking variable is unevenly distributed across the groups being pooled.
Also called: Simpson's reversal, Yule-Simpson effect.
The paradox is a vivid illustration of confounding: a treatment can look better within each severity group yet worse overall if the sicker patients were disproportionately given it. It warns that aggregated results and stratified results can point in opposite directions, and that neither is automatically correct. Which answer to trust depends on understanding the causal structure of the data, in particular whether the grouping variable is a confounder that should be adjusted for or a consequence of the exposure that should not.
Read the full Reading the Evidence blog.
This is a plain-language methodology definition for reading research. It is general education, not medical advice.