A problem that affects most longitudinal studies. For example, in a panel study, some of the original participants will not be participating at the end of the study—as a consequence of death, emigration, or failure to provide a forwarding address when moving house.