Journal of the Korean Data and Information Science Society

The Korean Data and Information Science Society (한국데이터정보과학회)
권호 표지

Percentile Envelope and Its Characteristic of Error Distribution for Supernormality

  • Published : 2001.10.30
Download PDF
⟨ Previous Next ⟩

Abstract

We introduce a new percentile envelope for diagnosing supernormality in regression analysis. Furthermore, we compare this percentile envelope, which is much simpler and easier, with Atkinson's and Flack and Flores' envelopes. Using percentile envelope, we investigate characteristics of normal probability plots with envelope for error distributions when supernormality is occurred. We give cautions that test result for normality assumption of errors can be reached the wrong conclusion by supernormality.

Keywords

References

  1. Biometrika v.68 Two graphical displays for outlying and influential observations in regression Atkinson, A. C.
  2. Plots, Transformations, and Regression Atkinson, A. C.
  3. An introduction to the bootstrap Efron, B.;Tibshirani, R. J.
  4. Technometrics v.31 Using simulated envelopes in the evaluation of normal probability plots of regression residuals Flack, V. F.;Flores, R. A.
  5. Methods for statistical data analysis of multivariate observations Gnanadesikan, R.
  6. Journal of the American Statistical Association v.82 Fine-tuning some resistant rules for outlier labeling Hoaglin, D. C.;Iglewicz, B.
  7. Journal of the American Statistical Association v.81 Performance of some resistant rules for outlier labeling Hoaglin, D. C.;Iglewicz, B.;Tukey, J. W.
  8. The Korean Communications in Statistics v.4 Plot Bimodality Lee, J.-Y.;Rhee, S.-W.
  9. Modern regression methods Ryan, T. P.
  10. Exploratory Data Analysis Turkey, J. W.
  11. Journal of the American Statistical Assoication v.75 Comment on "Some large sample tests for nonnormality in the linear regression model" Weisberg, S.; by White, H.(ed.);MacDonald, G. M.(ed.)
  12. Applied linear regression Weisberg, S.