论文信息 - Confidence region for parameters in replicated errors in variables model

Confidence region for parameters in replicated errors in variables model

COMPSTAT 2004 section: Multivariate analysis. Abstract: We consider replicated regression model with errors in variables (EIV model), also known as the measurement error model. The calibration problem is closely related to the considered model. In this model, we have derived the Kenward-Roger type of the confidence region for the unknown regression parameters, which is suggested to be used with small sample sizes. Our simulation study approved good statistical properties of the proposed confidence region.

[1] R. N. Kackar,et al. Approximations for Standard Errors of Estimators of Fixed and Random Effects in Mixed Linear Models , 1984 .

[2] L. Gleser. Measurement, Regression, and Calibration , 1996 .

[3] G. Casella,et al. Statistical Inference , 2003, Encyclopedia of Social Network Analysis and Mining.

[4] M. Kenward,et al. Small sample inference for fixed effects from restricted maximum likelihood. , 1997, Biometrics.

[5] Daniel R. Jeske,et al. Mean Squared Error of Estimation or Prediction under a General Linear Model , 1992 .

[6] H. Scheffé. A Statistical Theory of Calibration , 1973 .