A characterisation of a group such as the mean of a variable or an association between variables that changes for persons born at different times or places. In cross-sectional studies, this effect can give rise to inappropriate conclusions about the nature of developmental change. A well-worn example is finding that IQ appears to decline with age from early adulthood onward when, in fact, younger cohorts tend to have higher mean IQs than older ones as a consequence of better nutrition, schooling and the like (something known as the Flynn effect).
See Cohort, Cross-sectional design, Crystallised intelligence, Fluid intelligence, Flynn effect, Longitudinal design, Longitudinal studies, Multiple cohort study, Period effect