论文信息 - On the unreplicated ultrastructural model

On the unreplicated ultrastructural model

SUMMARY Dolby's (1976) ultrastructural model is investigated with emphasis on the case with no replications and one of the error variances known. It is shown that, in this situation, the maximum likelihood estimators are not consistent. The method of moments estimators are consistent and asymptotically normal and best in Fisher's sense within a broad class of asymptotically normal estimators discussed by Gleser (1983). Some connections with the errors-in-variables model are examined.

John W. Van Ness | Chi-Lun Cheng

[1] V. Barnett. A note on linear structural relationships when both residual variances are known. , 1967, Biometrika.

[2] M. W. Birch. A Note on the Maximum Likelihood Estimation of a Linear Structural Relationship , 1964 .

[3] M. Kendall,et al. The advanced theory of statistics , 1945 .

[4] A. Satorra,et al. Measurement Error Models , 1988 .

[5] A. Madansky. The fitting of straight lines when both variables are subject to error , 1959 .

[6] G. R. Dolby. The ultrastructural relation: A synthesis of the functional and structural relations , 1976 .

[7] D. V. Lindley,et al. Regression Lines and the Linear Functional Relationship , 1947 .

[8] The unreplicated ultrastructural relation: Large sample properties , 1978 .

[9] Leon Jay Gleser,et al. A note on G. R. Dolby's unreplicated ultrastructural model , 1985 .

[10] M. Solari. The “Maximum Likelihood Solution” of the Problem of Estimating a Linear Functional Relationship , 1969 .