Outlier Detection in Growth Curve Model

  • Published : 2003.05.31

Abstract

For the growth curve model with arbitrary covariance structure, known as unstructured covariance matrix, the problems of detecting outliers are discussed in this paper. In order to detect outliers in the growth curve model, the test statistics using U-distribution is established. After detecting outliers in growth curve model, we test homo and/or hetero-geneous covariance matrices using PSR Quasi-Bayes Criterion. For illustration, one numerical example is discussed, which compares between before and after outlier deleting.

Keywords

References

  1. 한국데이타정보과학회 2002년 추계학술발표 논문집 성장곡선모형의 이상점 탐지에 U-분포의 적용 심규박
  2. Sensitivity Analysis in Linear Regression Chatterjee, S.;Hadi, A.S.
  3. Detection of influential observations in linear regression Cook, R.D.;Weisberg, S.
  4. Journal of the American Statistical Association v.92 no.437 High-Breakdown linear discriminant analysis Hawkins, D.M.;McLachlan, G.J.
  5. Comm. Stat. Simulation Computer. B. v.9 Further tabulation of Hotelling's T² Davis, A.W.
  6. Journal of the American Statistical Association v.74 A predictive approach to model selection Geisser, S.;Eddy, W.F.
  7. Biometrics v.47 Detecting influential measurements in a growth curve model Liski, E.P.
  8. Biometrika v.61 F approximations to the distribution of Hotelling's T² McKeon, J.J.
  9. Biometrika v.51 A generalized multivariate analysis of variance model useful especially for growth curve problems Potthoff, R.F.;Roy, S.N.
  10. Journal of Multivariate Analysis v.31 Maximum likelihood estimates in multivariate linear model von Rosen, D.
  11. Multivariate Observations Seber, G.A.F.
  12. A study for the equality test of covariance matrices by PSR quasi-Bayes criterion in small sample Shim, K.B.
  13. Australian Journal of Statistics v.34 A modified groeth curve model and its applicationnn to clinical studies Weissfeld, L.A.;Kshirsagar, A.M.