DISCUSSION ON THE PAPER BY STEELE.

The article presents Michele Haynes' discussion of Fiona Steele's paper. It indicates how Steele presented a class of multilevel discrete-time event history that address some of the issues peculiar to the analysis of social process data and provided an thorough overview of how discrete-time event history models can be used to study the length of time before an individual experiences an event or a transition in state. It notes how Steele gave a comprehensive review of a range of multilevel, multiprocess models for analysing panel surveys.