New tests of heteroskedasticity in linear regression model

Lithuanian Mathematical Journal volume 47, pages248–265(2007)Cite this article

Abstract

In this paper, we present a class of tests for heteroskedasticity of various types in the linear regression model. These tests are based on the limit behavior of the polygonal process constructed from squared residuals. The law of test statistics under the null hypothesis is established, and the consistency is proved. By means of simulations these tests are compared with two classical tests (likelihood-ratio and Breusch-Pagan) for two types of heteroskedasticity (changed-segment and the type where the error variance is proportional to one of the components of the design matrix).

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

References

  1. 1.

    J. K. Binkley, Finite sample behavior of tests for grouped heteroscedasticity, The Review of Economics and Statistics, 74, 563–568 (1992).

    Article  Google Scholar 

  2. 2.

    T. S. Breusch and A. R. Pagan, A simple test for heteroscedasticity and random coefficient variation, Econometrica, 47, 1287–1294 (1979).

    MATH  Article  MathSciNet  Google Scholar 

  3. 3.

    J. M. Dufour, L. Khalaf, J. T. Bernard, and I. Genest, Simulation-based finite-sample tests for heteroskedasticity and ARCH effects, J. Econometrics, 122, 317–347 (2004).

    Article  MathSciNet  Google Scholar 

  4. 4.

    H. Glejser, A new test for heteroscedasticity, J. Amer. Statist. Assoc., 64, 316–323 (1969).

    Article  Google Scholar 

  5. 5.

    L. G. Godfrey, C. D. Orme, and J. M. C. Santos Silva, Simulation-based tests for heteroskedasticity in linear regression models: Some further results, Econometrics J., 9, 76–97 (2006).

    MATH  Article  MathSciNet  Google Scholar 

  6. 6.

    S. M. Goldfeld and R. Quandt, Some tests for heteroscedasticity, J. Amer. Statist. Assoc., 60, 539–547 (1965).

    Article  MathSciNet  Google Scholar 

  7. 7.

    D. Hamadouche, Principe d’invariance dans les espaces hölderiens pour des variables α-mé langeantes ou associées, Pub. IRMA Lille, 37-IX (1995).

  8. 8.

    D. Hamadouche, Weak convergence of smoothed empirical process in Hölder spaces, Statist. Probab. Lett., 36, 393–400 (1998).

    MATH  Article  MathSciNet  Google Scholar 

  9. 9.

    A. Račkauskas and Ch. Suquet, Necessary and sufficient condition for the functional central limit theorem in Hölder spaces, J. Theoret. Probab., 17, 221–243 (2004).

    Article  MathSciNet  Google Scholar 

  10. 10.

    A. Račkauskas and D. Zuokas, Off-line testing for a changed segment in the sample variance, Lith. Math. J., 45(2), 200–216 (2005).

    Article  Google Scholar 

  11. 11.

    J. Szroeter, A class of parametric tests for heteroscedasticity in linear econometric models, Econometrica, 46, 1311–1327 (1978).

    MATH  Article  MathSciNet  Google Scholar 

  12. 12.

    H. White, A heteroskedasticity-consistent covariance matrix and a direct test for heteroskedasticity, Econometrica, 48, 817–838 (1980).

    MATH  Article  MathSciNet  Google Scholar 

  13. 13.

    L. C. Young, General inequalities for Stieltjes integrals and the convergence of Fourier series, Math. Ann., 115, 581–612 (1938).

    MATH  Article  MathSciNet  Google Scholar 

Download references

Author information

Affiliations

  1. Vilnius University, Naugarduko 24, LT-03225, Vilnius, Lithuania

    A. Račkauskas & D. Zuokas

Authors
  1. A. Račkauskas
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. D. Zuokas
    View author publications

    You can also search for this author in PubMed Google Scholar

Additional information

Published in Lietuvos Matematikos Rinkinys, Vol. 47, No. 3, pp. 307–327, July–September, 2007.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Račkauskas, A., Zuokas, D. New tests of heteroskedasticity in linear regression model. Lith Math J 47, 248–265 (2007). https://doi.org/10.1007/s10986-007-0018-6

Download citation

Keywords

  • heteroskedasticity
  • FCLT
  • changed segment