An Actuarial Method for Assessing the Direction of Influence Between Two Datable Life Events

Abstract

A method is presented which can be used to assess the probability that one life event influences the chances of occurrence of a second event when the two events have overlapping periods of risk. The method takes into account the order of occurrence of the two events, corrects for the fact that persons who experience the first event have fewer years at risk of the second event than do persons without that experience, and allows holding precursors of both events constant, to detect spurious relationships. The method can be used for samples too small to permit using age-specific rates. Its use is illustrated with data from a study which attempts to answer the question "Does marijuana use lead to six outcomes: school dropout, arrest, jail, alcoholism, leaving the parental home early, and marrying early?"