scientific article; zbMATH DE number 1803731

From MaRDI portal
Publication:3147652





This second edition is aimed to give an updated introductory account of event history modeling techniques, with emphasis on the applications of continuous-time models. The main purpose of the book is threefold: (1) To demonstrate that event history models are an extremely useful approach to reveal causal relationships, or to map out a system of causal relations. (2) To introduce the reader to the computer program TDA (Transition Data Analysis). This program (written by G. Rohwer and U. Pötter, in 2000) estimates the sorts of models most frequently used with longitudinal data, in particular, discrete-time and continuous-time event history data. (3) The third goal pursued was to supplement the textbook by \textit{H.-P. Blossfeld, A. Hamerle} and \textit{K. U. Mayer}, Event history analysis. Statistical theory and application in the social sciences. (1989; Zbl 0725.62106), and to update the earlier edition of this book. Very briefly, the contents of the 10 enclosed chapters are the following: Chapter 1 (Introduction) discusses different observation plans and their consequences for causal modeling; fundamental concepts of event history analysis are surveyed. Chapter 2 (Event History Data Structures) describes event history data sets, their organization, and how to use such data sets with the computer program TDA. Chapter 3 (Nonparametric Descriptive Methods) examines basic nonparametric methods used to describe event history data, mainly life tables and Kaplan-Meier (product-limit) estimation methods. Chapter 4 (Exponential Transition Rate Models) shows how the simple model of the basic exponential transition rate can serve as an important starting point to all other transition rate models. Chapter 5 (Piecewise Constant Exponential Models) describes one of the most useful class of models for empirical research (enounced in the title), particularly helpful when time-dependent covariates can not be explicitly measured or the time-dependence of the process is not clear or not available. Chapter 6 (Exponential Models with Time-Dependent Covariates) discusses time-dependent covariates in transition rate models, considered to be the most important step forward in event history analysis with respect to the empirical study of social change. The examples are restricted to exponential and piecewise exponential models. In particular, the problem of how to model parallel and inter-dependent processes is introduced. Chapter 7 (Parametric Models of Time-Dependence) considers a variety of models with parametrically specified duration-dependent transition rate, especially Gompertz-Makeham, Weibull, log-logistic, log-normal, and sickle models. Chapter 8 (Methods to Check Parametric Assumptions) investigates the question of goodness-of-fit for parametric transition rate models. Simple graphical checks based on transformed survivor functions and generalized residuals are described. Chapter 9 (Semi-Parametric Transition Rate Models) introduces an alternative (proposed by D. R. Cox in 1972) to specific parametric models, which is able to specify only a functional form for the influence of covariates, but leaving the shape of the transition rate as unspecified as possible. Finally, Chapter 10 (Problems of Model Specification) takes into account problems of model specification and, in particular, transition rate models with unobserved heterogeneity. The discussion is mainly critical and the examples are restricted to using a gamma mixing distribution. The book and its contents addresses a large spectrum of readers and application fields, from students to researchers, and from economics, social sciences, psychology, and political sciences to medical and epidemiology applications, communication analysis, etc.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3147652)