Skip to main content
Log in

On null and alternative distribution of statistic of two-side discordancy test for an extreme outlier

  • Published:
Russian Mathematics Aims and scope Submit manuscript

Abstract

We find the joint distribution of Grubbs statistics for a normal sample. Those statistics are standardized maximum and standardized minimum. We note some properties of the joint distribution function. We apply the joint distribution function and find the exact distribution of the test statistic which uses in two-sided discordancy test for an extreme outlier. We obtain recursive relationships for the distribution function of the statistic, which uses in two-sided discordancy test. We obtain the region of critical values of the statistic, where the significance level of criteria equals the double significance level of the Grubbs criteria. We apply the joint distribution of Grubbs statistics and find the power function for the criteria in the case of a normal sample with a single outlier.

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

  1. David, H. A. Order Statistics (John Wiley & Sons, 1970; Nauka, Moscow, 1979) [in Russian].

    MATH  Google Scholar 

  2. Barnett, V., Lewis, T. Outliers in Statistical Data (John Wiley & Sons, Chichester, 1984).

    MATH  Google Scholar 

  3. Grubbs, F. “Sample Criteria for Testing Outlying Observations,” Ann. Math. Statist. 21, No. 1, 27–58 (1950).

    Article  MATH  MathSciNet  Google Scholar 

  4. Zhang J., Keming Y. “The Null Distribution of the Likelihood-Ratio Test for One or Two Outliers in a Normal Sample,” An Official Journal of the Spanish Society of Statistics and Operations Research 15, No. 1, 141–150 (2006).

    MATH  Google Scholar 

  5. Bol’shev, L. N., Smirnov, N. V. Tables of Mathematical Statistics (Nauka, Moscow 1983) [in Russian].

    MATH  Google Scholar 

  6. Kazakavichus, K. A. “Approximate Formulas for Statistical Processing Results of Mechanical Tests,” Zavodskaja Laboratoriya 54, No. 12, 82–85 (1988).

    Google Scholar 

  7. Verma, S. P., Quiroz-Ruiz, A. “Critical Values for 22 Discordancy Test Variants for Outliers in Normal Samples up to Size 100, and Application in Science and Engineering,” Revista Mexicana de Ciencias Geologicas 23(3), 302–319 (2006).

    Google Scholar 

  8. Livesey, J.H. “Curtosis Provides a Good Omnibus Test for Outliers in Small Samples,” Clinical Biochemistry 40, 1032–1036 (2007).

    Article  Google Scholar 

  9. Shiryaeva, L. K. “Calculation of Measures of Power of the Grubbs Criterion for a Test for an Outlier,” Sib. Zh. Ind. Mat. 13, No. 4, 141–154 (2010).

    MATH  MathSciNet  Google Scholar 

  10. Shiryaeva, L. K. “UsingHermitian Functions for the Investigations of Power Properties of Grubbs Criterion,” Vestn. Samarsk. Gos. Tekhn. Univ. Ser. Fiz.-matem. nauki 29, No. 4, 131–145 (2012) (in Russian).

    Article  Google Scholar 

  11. Regions of Russia. Social-Economic Indices. 2012. Statistical Collection (Rosstat, Moscow, 2012) [in Russian].

  12. Zhang J., Wang X. “Unmasking Test for Multiple Upper or Lower Outliers in Normal Samples,” J. Appl. Statistics 25, No. 2, 257–261 (1998).

    Article  MATH  Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to L. K. Shiryaeva.

Additional information

Original Russian Text © L.K. Shiryaeva, 2014, published in Izvestiya Vysshikh Uchebnykh Zavedenii. Matematika, 2014, No. 10, pp. 62–78.

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Shiryaeva, L.K. On null and alternative distribution of statistic of two-side discordancy test for an extreme outlier. Russ Math. 58, 52–66 (2014). https://doi.org/10.3103/S1066369X14100089

Download citation

  • Received:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.3103/S1066369X14100089

Keywords

Navigation