This project studies determinants of entry into and exit from family-related adult roles, including marital and premarital child birth, marriage, separation, and cohabitation. It goes beyond previous methodology by distinguishing between deter- minants of the timing of events and determinants of the lifetime probability of the events. Previous methods cannot distinguish timing and probability, so, for example, they cannot tell whether a higher divorce rate means members in the population are generally more likely to experience divorce, or whether they just experience it earlier than members of some other population. The PI has developed new methods and models to see, for example, whether certain risk factors associated with high divorce rates cause higher probabil- ities or accelerate the timing of divorce for those couples who will eventually divorce. Software developed for this project will be available for use by other investigators using similar large data sets.
