Longitudinal models for assessing cognitive development in early childhood

We develop likelihood-based statistical inference to assess change in performance of young children on repeated executive function tests. The data comprise sequences of non-independent binary outcomes that are taken at a pre-specified set of follow-up times. In this talk we address some methodological questions regarding the two executive functions inhibitory control and working memory. To do this we build random-effect-transition regression models that take into account variability within individuals, dependence between results on successive testing occasions and inter-relationships between tests.