6. Maximum Likelihood (ML) and Restricted Maximum Likelihood (REML)

  1. Shayle R. Searle,
  2. George Casella and
  3. Charles E. McCulloch

Published Online: 27 MAY 2008

DOI: 10.1002/9780470316856.ch6

Variance Components

Variance Components

How to Cite

Searle, S. R., Casella, G. and McCulloch, C. E. (2008) Maximum Likelihood (ML) and Restricted Maximum Likelihood (REML), in Variance Components, John Wiley & Sons, Inc., Hoboken, NJ, USA. doi: 10.1002/9780470316856.ch6

Publication History

  1. Published Online: 27 MAY 2008
  2. Published Print: 13 MAR 1992

ISBN Information

Print ISBN: 9780471621621

Online ISBN: 9780470316856

SEARCH

Keywords:

  • probability distribution;
  • random effects;
  • variance components;
  • residual errors;
  • residual maximum likelihood

Summary

The prelims comprise:

  • The model and likelihood function

  • The ML estimation equations

  • Asymptotic dispersion matrices for ML estimators

  • Some remarks on computing

  • ML results for 2-way crossed classification, balanced data a. 2-way crossed, random model, with interaction

  • Restricted maximum likelihood (REML)

  • Estimating fixed effects in mixed models

  • ML or REML?

  • Summary

  • Exercises