Abstract
Multiple-spell discrete-time survival analysis can be used to investigate the repeated occurrence of a single event, or the sequential occurrence of disparate events, including: students’ and teachers’ entries into, and exits from, school; childrens’ progress through stages of cognitive reasoning; disturbed adolescents’ repeated suicide attempts; and so forth. In this article, we introduce and illustrate the method using longitudinal data on exit from, and reentry into, the teaching profession. The advantages of the approach include: (a) applicability to many educational problems; (b) easy inclusion of time-invariant and time-varying predictors; (c) minimal assumptions—no proportional-hazards assumption is invoked and so the effects of predictors can vary over time within, and across, spells; and (d) all statistical models can be fit with a standard logistic regression package.
Get full access to this article
View all access options for this article.
