Local Influence Analysis of the Equicorrelation Model

  • Kim, Myung-Geun (Department of Applied Statistics, Seowon University) ;
  • Jung, Kang-Mo (Department of Information & Statistics, Kunsan National University)
  • 발행 : 2002.12.01

초록

The influence of observations in the equicorrelation model is investigated using the local influence approach when all parameters or subsets of parameters are of interest. When a parameter of interest is scalar, an analytical form of the local influence measure can be found. We will derive a measure for identifying observations that have a large influence on the test of fitting the equicorrelation model. An example is given for illustration.

키워드

참고문헌

  1. Journal of the Royal Statistical Society v.C36 A new graphical method for detecting single and multiple outliers in multivariate data Bacon-Shone, J.;Fung, W. K.
  2. Outliers in Statistical Data(3rd ed.) Barnett, V.;Lewis, T.
  3. Journal of the Royal Statistical Society v.B48 Assessment of local influence (with discussions) Cook, R. D.
  4. Biometrics v.52 Case influence on the intraclass correlation coefficient estimate Giraudeau, B.;Mallet, A.;Chastang, C.
  5. Applied Multivariate Statistical Analysis(3rd ed.) Johnson, A. J.;Wichern, D. W.
  6. Computational Statistics & Data Analysis v.24 Local influence in maximum likelihood factor analysis Jung, K. -M.;Kim, M. G.;Kim, B. C.
  7. Communications in Statistics-Theory and Methods v.24 Local influence in multivariate regression Kim, M. G.
  8. Journal of the American Statistical Association v.83 Regression transformation diagnostics using local influence Lawrance, A. J.
  9. Methods of Multivariate Analysis Rencher, A. C.