2陈希镇.The Deviation between the Least Squares and the Best Linear Unbiased Estimation of the Mean Vector in the Linear Mode[J].华东师范大学学报：自然科学版,2001,:04-04.
3王松桂.线性回归系数的一种新估计[J].中国科学：A辑,1988,10:1033-1040.
4Haberman, S.J. How much do Gauss-Markov and least squares estimaters differ a coordinate free approach[J]. Ann. Statist, 1975,3:982-999.
5Rao, C. R. The inefficiency of least squares extrnsion of the Rantorovich ineqttality[J]. Linear Algebra Appl. , 1985,70:249-255.
6Khati,C. G. ,Rao, C.R. Some generalizations of kantorovich inequality[J]. Sanrhya Sera, 1982, 44:91-102.
7Fujikoshi, Y. and Rao, C.R., Selection of covariables in the growth curve model, Biometrika, 78(1991),779-785.
8Wang S.G. (王松桂) Liski E.P. Nummi T.Two-way selection of covaxiables in multivariate growth curve models[J].Linear Algebra and its Applications,(1999):333-342.
9Cook, R.D., Detection of influential observation in linear regression, Technomitrics, 19(1977), 15-18.
