ricerca
avanzata

Antedependence Models for Longitudinal Data - 9781420064261

Un libro in lingua di Zimmerman Dale L., Nunez anton Vicente A. edito da Taylor & Francis, 2009

  • € 233.80
  • Il prezzo è variabile in funzione del cambio della valuta d’origine

Zimmerman (U. of Iowa, US) and Núñez-Antón (The U. of the Basque Country, Spain) describe a class of models for longitudinal data know as antedependence models or transition models, which postulate that certain conditional independencies exist among the observations that are related to their time ordering. They are particularly useful for modeling longitudinal data that exhibit serial correlation (correlation that decays as the elapsed time between measurements increases). The authors describe the models and their properties, focusing on unstructured and structured antedependence individually, and then present inference procedures for the models in chapters that cover informal model identification via simple summary statistics and graphical methods, maximum likelihood and residual maximum likelihood estimation of parameters, likelihood ratio tests and penalized likelihood model selection criteria for the model's covariance structure, and mean structure. They use illustrative examples throughout and, in a later chapter, use these examples to compare the performance of antedependence models to other models commonly used for longitudinal data. Annotation ©2009 Book News, Inc., Portland, OR (booknews.com)

Informazioni bibliografiche