The Use of SPlus Software to Analyse Event History Data: An Application to the Early Career Promotion of Nurses in UK

摘要

This paper discusses the benefits of using SPlus software to carry out longitudinal analyses such as event history methods. An illustration is provided using three-year career histories of a nationally drawn cohort of recently qualified registered nurses, who completed a diploma programme in the UK between August 1997 and September 1998. For the purposes of this paper, the time to first promotion is examined using a Cox proportional hazards model with a frailty term included, to allow for unobserved heterogeneity. The results indicated that men have a slightly higher rate of promotion compared with women, and particularly when compared with women who took a maternity break during their early career. Age is also a significant factor, with younger cohort members being promoted at a higher rate than their older colleagues. The inclusion of a frailty term to account for clustering at the university level also significantly improved the model. It is shown that SPlus, which is a powerful, user-friendly statistical package, can be easily utilized by researchers to carry out event history analysis as well produce high quality cumulative hazard and diagnostic plots using its sophisticated graphics facilities.