Coping with complex individual histories: Comparing life course methods with an application to partnership transitions in Norway

As variation in the pattern of family life courses has increased over the past 50 years, the techniques available to analyse life course data have also expanded. While event history analysis is commonly applied, this is not always suitable, and more holistic approaches such as sequence analysis have been proposed as alternatives. As research tends to be interested in explaining more complexity in the family life course, it is necessary to extend our methodological toolkit by increasing the complexity of event history models (multistate event history models and simultaneous event history models) or applying other promising methods, such as sequence analysis and latent class growth models. The aim of this paper is to compare and contrast simultaneous event history models, multistate models, sequence analysis, and latent class growth curve models to studying the family life course. The advantages and weaknesses of each of these methods are highlighted by applying them to the same empirical problem. Using data from the first wave of the Norwegian Generations and Gender Survey from 2007/2008 for women in birth cohorts 1945-1954, 1955-1964, and 1965-1974, we model changes in partnership status across the life course, with education as the primary covariate of interest.