Constant Error Variance Assumption in Random Effects Linear Model

  • Published : 1995.12.01

Abstract

When heteroscedasticity occurs in random effects linear model, the error variance may depend on the values of one or more of the explanatory variables or on other relevant quantities such as time or spatial ordering. In this paper we derive a score test as a diagnostic tool for detecting non-constant error variance in random effefts linear model based on the model expansion on error variance. This score test is compared to loglikelihood ratio test.

Keywords

References

  1. Journal of Korean Statistical Society v.19 no.2 Diagnostics for Heteroscedasticity in Mixed Linear Models Ahn,Chul H.
  2. Journal of Royal Statistical Society, Ser. A v.143 Sampling and Bayes inferencein Scientific Modelling and robustness(with discussion) Box,G.E.P.
  3. Journal of American Statistical Association v.78 Score tests for regression models Chen,C.
  4. Residuals and Influence in Regression Cook,R.D.;Weisberg,S.
  5. Biometrika v.70 Diagnostics for heteroscedasticity in Regression Cook,R.D.;Weisberg,S.
  6. Journal of Royal Statistical Society, Ser. B v.48 Assessment of Local Influence(with discussion) Cook,R.D.
  7. Technometrics v.29 Diagnostics for Mixed-Model Analysis of Variance Cook,R.D.;Beckman,R.;Nachtsheim,C.
  8. Theoretical Statistics Cox,D.R.;Hinkley,D.V.
  9. Journal of American Statistical Association v.80 Weighted normal plot Dempster,A.P.;Ryan,L.M.
  10. Kronecker Products and Matrix Calculus with Application Graham,A.
  11. Biometrika v.72 Transformation diagnostics for linear model Hinkley,D.V.
  12. Proceedings of International Biometrics Conference Diagnostic methods in variance component estimation Hocking,R.R.
  13. Matrix Derivatives Rogers,G.S.
  14. Journal of Royal Statistical Society, Ser. B v.51 Generalized linear models with varying dispersion Smyth,Gordon