Linear contrasts in experimental design with non-identical error distributions

Senoglu B., Tiku M.

BIOMETRICAL JOURNAL, vol.44, no.3, pp.359-374, 2002 (SCI-Expanded) identifier

  • Publication Type: Article / Article
  • Volume: 44 Issue: 3
  • Publication Date: 2002
  • Doi Number: 10.1002/1521-4036(200204)44:3<359::aid-bimj359>3.0.co;2-k
  • Journal Name: BIOMETRICAL JOURNAL
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.359-374
  • Keywords: experimental design, non-normality, generalized logistic, modified likelihood, linear contrast, non-identical blocks
  • Middle East Technical University Affiliated: No

Abstract

Estimation of linear contrasts in experimental design, and testing their assumed values, is considered when the error distributions from block to block are not necessarily identical. The normal-theory solutions are shown to have low efficiencies as compared to the solutions presented here.