Evidence-appraisal glossary
Collider bias
Collider bias is a distortion that appears when you select or adjust for a variable that two other factors both affect. Filtering on that shared effect can create a fake link, or hide a real one, between factors that are not actually related in the wider population.
Also called: Collider stratification bias, Conditioning on a collider, Berkson's bias.
What it is. A collider is a variable influenced by two or more other variables. When a study conditions on a collider, by restricting the sample, stratifying, matching, or statistically adjusting for it, it opens a backdoor path that can manufacture an association between those causes even if none exists. A famous case is Berkson's bias in hospital-based studies, where sampling only admitted patients links diseases that are unrelated in the general population.
How to use it when reading a study. Ask what defined who entered the sample and which variables were adjusted for. If the selection criterion or a control variable is plausibly a common effect of both the exposure and the outcome (or of things related to them), treat any reported association cautiously. Watch for it in case-control designs, studies of survivors, loss to follow-up, and analyses that adjust for mediators or post-exposure variables. Unlike confounding, adjusting harder here makes the bias worse, not better.
This is a plain-language methodology definition for reading research. It is general education, not medical advice.