Chapter 9. Alternative Optimal Designs for Linear Models

  1. Martijn P. F. Berger1 and
  2. Weng Kee Wong2

Published Online: 27 MAY 2009

DOI: 10.1002/9780470746912.ch9

An Introduction to Optimal Designs for Social and Biomedical Research

An Introduction to Optimal Designs for Social and Biomedical Research

How to Cite

Berger, M. P. F. and Wong, W. K. (2009) Alternative Optimal Designs for Linear Models, in An Introduction to Optimal Designs for Social and Biomedical Research, John Wiley & Sons, Ltd, Chichester, UK. doi: 10.1002/9780470746912.ch9

Author Information

  1. 1

    Department of Methodology and Statistics, Maastricht University, The Netherlands

  2. 2

    Department of Biostatistics, School of Public Health, University of California, Los Angeles, USA

Publication History

  1. Published Online: 27 MAY 2009
  2. Published Print: 29 MAY 2009

ISBN Information

Print ISBN: 9780470694503

Online ISBN: 9780470746912

SEARCH

Keywords:

  • alternative optimal designs for linear models;
  • D-optimal design minimizing volume of confidence ellipsoid for the model parameters;
  • DA, indicating dependence of the determinant on the matrix A;
  • Bayesian framework for designing a study, including its advantages and potential difficulties;
  • maximin optimal designs, operating in a similar way as minimax optimal designs;
  • heteroscedastic G-optimal designs;
  • minimax design strategy - simpler to implement in practice;
  • multiple-objective optimal designs, for addressing important concerns arising from model assumptions

Summary

This chapter contains sections titled:

  • Introduction

  • Information matrix

  • DA- or Ds-optimal designs

  • Extrapolation optimal design

  • L-optimal designs

  • Bayesian optimal designs

  • Minimax optimal design

  • Multiple-objective optimal designs

  • Summary