Modelling ordinal recurrent events.

An ordinal outcome is one which comprises a range of categories that are inherently ordered in some sense. An example of such an outcome is degree of supervision required (from ‘not at all’ to ‘very closely’) in the current job of an individual. In addition, an individual's outcome may be recorded repeatedly over time. For example, information may be available on the level of supervision required in each and every job in which an individual had been employed. These ordinal outcomes, measured repeatedly over time, are defined to be ordinal recurrent events. The early stages of a project seeking to develop and implement statistical models for these ordinal recurrent events will be described. Suggestions for future progress of the project, including possible solutions to problems already anticipated, will be discussed