A class of invariant consistent tests for multivariate normality (Q3978078)

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 13681
Language Label Description Also known as
default for all languages
No label defined
    English
    A class of invariant consistent tests for multivariate normality
    scientific article; zbMATH DE number 13681

      Statements

      A class of invariant consistent tests for multivariate normality (English)
      0 references
      0 references
      0 references
      25 June 1992
      0 references
      affine invariant tests
      0 references
      composite hypotheses
      0 references
      empirical characteristic function
      0 references
      scaled residuals
      0 references
      limiting null distribution
      0 references
      power performance
      0 references
      Monte Carlo study
      0 references
      test for multivariate normality
      0 references
      0 references
      0 references
      Let \(X_ 1,\dots,X_ n\) be independent, identically distributed random vectors in \(R^ d\), with sample mean \(\overline{X}_ n\) and sample covariance matrix \(S_ n\). The authors propose a class of affine invariant tests for the composite hypotheses \(H_ d\) : the law of \(X_ 1\) is a non-degenerate normal distribution, which are consistent against any fixed non-normal alternative distribution. The test statistic is a weighted integral of the squared modulus of the difference between the empirical characteristic function of the scaled residuals \(S_ n^{- 1/2}(X_ j-\overline{X}_ n)\) and its pointwise limit \(\exp(-2^{- 1}\| t\|^ 2)\) under \(H_ d\). The weight function is taken to be \((2\pi\beta^ 2)^{-d/2}\exp(-\| t\|^ 2/2\beta^ 2)\) for the varying parameter \(\beta\). The limiting null distribution of the test statistic is obtained. Power performance of the new tests is assessed in a Monte Carlo study. It is shown that the choice \(\beta=0.5\) yields a powerful test for multivariate normality against alternative distributions having heavy tails.
      0 references

      Identifiers