Longitudinal categorical data analysis (Q476361)

This book focuses on the analysis of complete sets of longitudinal categorical data consisting of responses and covariates collected repeatedly from the same individuals over a small period of time. For the analysis, several various parametric correlation models for repeated multinomial responses are developed by generalising correlation models for longitudinal binary data, using, specifically, dynamic models to relate repeated multinomial responses. Starting with an overview of regression models for cross-sectional univariate categorical data, it then moves on the presentation of regression models specifically for longitudinal stationary and non-stationary categorical data, and multinomial models for cross-sectional and longitudinal bivariate categorical data. The book is mainly addressed to graduate students and researchers in the fields of statistics, social sciences and other areas of applied statistics, although some chapters are passable to undergraduate students of senior statistic courses as well.