The extent to which the results of a study accurately represent the causal relationship between an intervention and an outcome in the particular circumstances of that study.