No label defined (Q3217448)

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 3883420
Language Label Description Also known as
default for all languages
No label defined
    English
    No label defined
    scientific article; zbMATH DE number 3883420

      Statements

      0 references
      1984
      0 references
      goodness of fit
      0 references
      sample spacings
      0 references
      higher order spacings
      0 references
      order statistics
      0 references
      m-step spacings
      0 references
      generalizations of the Greenwood statistic
      0 references
      Let \(X'_ 1,...,X'_{n-1}\) be the order statistics obtained from i.i.d. random variables \(X_ 1,...,X_{n-1}\) with distribution function F(x) on [0,1]. Let \(\{D_ i^{(m)}\}\) be the overlapping m-step spacings defined by \(D_ i^{(m)}=X'_{i+m}-X'_ i\), \(i=0,...,n-1\), and let \(\{D_{im}^{(m)}\}\) be the non-overlapping m-step spacings defined by \(D_{im}^{(m)}=X'_{(i+1)m}-X'_{im},\) \(i=0,1,...,[n/m]-1\), where [n/m] denotes the integer part of (n/m). Let \[ T_{2n}(m)=n^{- 1}\sum^{n-1}_{i=0}h(nD_{im}^{(m)})\quad and\quad T_{3n}(m)=n^{-1}\sum^{n-1}_{i=0}h(nD_ i^{(m)}). \] \(T_{2n}(m)\) and \(T_{3n}(m)\) are generalizations of the Greenwood statistic \(G_{1n}=n^{-1}\sum^{n-1}_{i=0}(nD_ i)^ 2.\) The statistic \(G_{1n}\) is known to be locally most powerful among all tests based symmetrically on \(G_{1n}.\) In this paper, it is shown that \(T_{2n}(m)\) and \(T_{3n}(m)\) are asymptotically efficient. Further, it is also shown that the limiting efficiency increases if the test is based on \(T_{3n}(m)\) as opposed to the test based on \(T_{2n}(m)\). Even though the result of this paper suggests larger m values are always better, the choice of m is an open question.
      0 references

      Identifiers