Skip to main content
Log in

On the reliability of the chi-square test

  • Publications
  • Published:
Metrika Aims and scope Submit manuscript

Summary

The classical χ2 test of goodness of fit depends on the partition chosen to calculate it. In this work the covariance structure of the family of all χ2 tests obtained by translation (modulo 1) of a uniform partition of the interval [0, 1) it is determined, under the hypothesis of uniform distribution on [0, 1). This is used to define and compute an index which measure the indeterminacy of the χ2 test due to the arbitrary choice of the partition to compute it.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others

References

  • Beran RJ (1969) Asymptotic theory of a class of tests for uniformity of a circular distribution. Ann Math Statist 40:1196–1206

    Google Scholar 

  • Bickel PJ, Rosenblatt M (1973) On some global measures of the deviations of density function estimates. Ann Statist 1:1071–1096

    Google Scholar 

  • Cochran WG (1952) The χ2 test of goodness of fit. Ann Math Statist 23:315–345

    Google Scholar 

  • Gumbell EJ (1943) On the reliability of the classical chi square test. Ann Math Statist 14:253–263

    Google Scholar 

  • Kendall MG, Stuart A (1958) The advanced theory of statistics, vol 1

  • Wichura MJ (1971) A note on the weak convergence of stochastic processes. Ann Math Statist 42:1969–1972

    Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Rights and permissions

Reprints and permissions

About this article

Cite this article

Viollaz, A.J. On the reliability of the chi-square test. Metrika 33, 135–142 (1986). https://doi.org/10.1007/BF01894740

Download citation

  • Received:

  • Issue Date:

  • DOI: https://doi.org/10.1007/BF01894740

Keywords

Navigation