Experimental designs with dependent observations: New results and comments on previous results (Q4725547)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 3999042
Language Label Description Also known as
default for all languages
No label defined
    English
    Experimental designs with dependent observations: New results and comments on previous results
    scientific article; zbMATH DE number 3999042

      Statements

      Experimental designs with dependent observations: New results and comments on previous results (English)
      0 references
      1987
      0 references
      factorial ANOVA models
      0 references
      dependent observations
      0 references
      covariance structures
      0 references
      Bartlett test
      0 references
      chi square distribution
      0 references
      orthogonal decomposition of the total sum of squares
      0 references
      F-statistics
      0 references
      Fisher F-random variables
      0 references
      covariance matrix
      0 references
      This paper deals with those ANOVA models in which the observations are normal and dependent. It is shown that for such models with orthogonal decomposition of the total sum of squares, the F-statistics for testing statistical hypotheses are distributed proportionally to Fisher F-random variables if the covariance matrix of the observations has a given explicit form. The following is the main result of this paper: Result: If \(X=(X_ 1,X_ 2,...,X_ n)'\sim N(\mu,\Sigma)\), \(\mu\) being the mean vector and \(\Sigma\) the covariance matrix, if the \(n\times n\) real symmetric matrices \(S_ 0,S_ 1,...,S_ k\) are such that \(S^ 2_ i=S_ i\) for each i, \(S_ iS_ j=0\), \(i\neq j\), \(S_ 0+S_ 1+...+S_ k=I-n^{-1}U=S\) (say), (I being the identity matrix, U the matrix with each element equal to one), and if \(\alpha_ i\) \((i=0,1,...,K)\) are positive constants, then a necessary and sufficient condition for \(X'S_ iX\sim \alpha_ i\chi^ 2(v_ i,\alpha_ i^{- 1}\mu 'S_ i\mu)\), \(X'S_ iX\) is independent of \(X'S_ jX\) (i\(\neq j)\) \((v_ i\) being the rank of \(S_ i)\) is that \(\Sigma =\sum^{K}_{i=0}\alpha_ iS_ i+\Gamma +\Gamma '\) where \(\Gamma\) is a matrix with equal rows having \(\gamma_ 1,\gamma_ 2,...,\gamma_ n\) as elements.
      0 references

      Identifiers