[1]Gross J, Trenkler G,Liski E P. Necessary and sufficient conditions for superiority of misspecified restricted least squares regression estimator. J Statist Planning Inference， 1998, 71：109-116

[2]Razzaghi M. A superiority problem in misspecified restricted linear model. Commu n Statist Ser B, 1987,16:899-902

[3]Kabe D G, Gupta A K. On a superiority problem in misspecified restricted linear models. Commun Statist Ser A, 1989, 18: 1753-1757

[4]Zhang B X, Liu B S. Some further remarks on a superiority problem in misspecified restricted linear models. Metrika, 2000,52(2):173-181

[5]Rao C R. Unified theory of linear estimation. Sankhya Ser A,1971,33: 371-394

[6]Aitken A C. On least squares and linear combination of observations. Proc Roy Soc Edinburgh Sect, 1935,A55:42-49

[7]Rao C R. Linear Statistical Inference and Its Applications.  2nd ed, Ne w York:  Wiley,  1973

[8]Albert A. The GaussMarkov theorem for regression models with possibly singular covariances. SIAM Appl Math, 1973,24: 182-187

[9]Searle S R. Extending some results, proofs for the singularlinear model. Linear Algebra Appl, 1994,210: 139-151

[10]Bhimasankaram P, Sengupta D. The linear zero functions approach  to linear models. Sankhya Ser B, 1996,58: 338-351

[11]Jammalamadaka S R, Sengupta D. Changes in the general linear model: A un ified approach Linear Algebra Appl, 1999,289: 225-242

[12]Puntanen S, Scott A J. Some further remarks on the singular linear model. Linear Algebra Appl,  1996,237/238: 313-327

[13]Horn R A, Johnson C R. Matrix Analysis. Cambridge： Cambridge University  Press，1985

[14]Baksalary J K, Kala R, Klaczynski K. The matrix inequality M≥B^*MB . Linear Algebra  Appl, 1983, 54: 77-86