Longitudinal Study

A longitudinal study collects data from the same participants at two or more points in time, allowing genuine change to be observed rather than inferred. Panel designs follow the same individuals; cohort designs follow a group sharing a defining characteristic.

The methodological payoff is temporal precedence. Because the predictor is measured before the outcome, longitudinal data supports causal claims far more strongly than a cross-sectional snapshot, which is why journals reviewing mediation models increasingly ask for it.

The costs are time, expense and attrition — participants drop out between waves, and if dropout is systematic rather than random it biases the results. Report your retention rate, compare completers with dropouts on baseline variables, and state how missing data across waves was handled.