Dynamic Mixed Models for Familial Longitudinal Data

摘要

Familial data arise when responses, and possibly accompanying covariates, are available on independent clusters of cases, such as families, where the relatedness of the individuals within the clusters generates correlations among the responses. Longitudinal data arise when repeated measurements are taken on independent individuals across time. Correlations in familial data might be incorporated via a random cluster-specific effect. Those in longitudinal data might be modelled within a single individual via autocorrelation structures familiar in time-series models, for example. In this book the author considers appropriate models for these different types of correlated data and provides comprehensive inferential and numeric procedures to accompany them, particularly for discrete count and binary data.