A semiparametric multilevel survival model

Authors: Zhang W.¹; Steele F.²

Source: Journal of the Royal Statistical Society: Series C (Applied Statistics), Volume 53, Number 2, April 2004 , pp. 387-404(18)

Publisher: Wiley-Blackwell

Buy & download fulltext article:

Price: $48.00 plus tax (Refund Policy)

Abstract:

Summary.

We propose a semiparametric multilevel survival model for clustered duration data in which the effect of a continuous covariate is represented by an unspecified, possibly non-linear, function. This model makes no distributional assumption about the cluster level random effects. The performance of the method is assessed via Monte Carlo simulations. The model is applied in an analysis of first-birth intervals in Bangladesh to examine period effects in the timing of first births, while allowing for clustering within communities; the analysis reveals a non-linear trend in the first-birth interval over time.

Keywords: Clustered data; First-birth interval; Multilevel model; Semiparametric survival model

Document Type: Research article

DOI: http://dx.doi.org/10.1111/j.1467-9876.2003.05056.x

Affiliations: 1: University of Kent, Canterbury, UK. 2: Institute of Education, London, UK

Publication date: 2004-04-01