Chapter 5. Alternative Strategies for Model Selection and Inference Using Information-Theoretic Criteria

  1. Howard B. Stauffer

Published Online: 11 APR 2007

DOI: 10.1002/9780470185094.ch5

Contemporary Bayesian and Frequentist Statistical Research Methods for Natural Resource Scientists

Contemporary Bayesian and Frequentist Statistical Research Methods for Natural Resource Scientists

How to Cite

Stauffer, H. B. (2007) Alternative Strategies for Model Selection and Inference Using Information-Theoretic Criteria, in Contemporary Bayesian and Frequentist Statistical Research Methods for Natural Resource Scientists, John Wiley & Sons, Inc., Hoboken, NJ, USA. doi: 10.1002/9780470185094.ch5

Author Information

  1. Mathematics Department, Humboldt State University, Arcata, California, USA

Publication History

  1. Published Online: 11 APR 2007
  2. Published Print: 16 NOV 2007

ISBN Information

Print ISBN: 9780470165041

Online ISBN: 9780470185094

SEARCH

Keywords:

  • statistical modeling and multiple linear regression;
  • Markov Chain Monte Carlo (MCMC) simulation;
  • frequentist statistical analysis

Summary

This chapter contains sections titled:

  • Alternative Strategies for Model Selection and Inference: Descriptive and Predictive Model Selection

  • Descriptive Model Selection: A Posteriori Exploratory Model Selection and Inference

  • Predictive Model Selection: A Priori Parsimonious Model Selection and Inference Using Information-Theoretic Criteria

  • Methods of Fit

  • Evaluation of Fit: Goodness of Fit

  • Model Averaging

  • Applications: Frequentist Statistical Analysis in S-Plus and R; Bayesian Statistical Analysis in WinBUGS

  • Summary

  • Problems