Using permutation tests to reduce type I and II errors for small ruminant research


Oender H.

JOURNAL OF APPLIED ANIMAL RESEARCH, vol.32, no.1, pp.69-72, 2007 (SCI-Expanded, Scopus)

  • Publication Type: Article / Article
  • Volume: 32 Issue: 1
  • Publication Date: 2007
  • Journal Name: JOURNAL OF APPLIED ANIMAL RESEARCH
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.69-72
  • Keywords: permutation test, ANOVA, power of the test, type I and II errors, hair length, MODELS
  • Ondokuz Mayıs University Affiliated: Yes

Abstract

Although parametric tests (F or t) are considerably effective, these are sometimes ineffective when the assumptions needed by model are not provided. In such a case, permutation test unaffected by the assumptions can be applied as a non-parametric method. It has been observed by citing an example that permutation test produces more reliable results than one-way ANOVA in terms of type I error rate and power of the test and permutation test is recommended in order to avoid type I and II errors and to prevent the potential profit lost.