Journal of the American Statistical Association volume 80, issue 391, P698-703 1985 DOI: 10.1080/01621459.1985.10478171 View full text
|
|
Share
A. S. Hedayat, Dibyen Majumdar

Abstract: Consider two experiments with homoscedastic models E( Yi) = X i 8 + Z i y , V(Yi) = d I n i ( i = 1, 2). The matrix di = X ; ( Z , -Zi(ZIZi)-Zt')Xi can be called the information matrix of linear estimable functions of B for experiment i . The information matrix 9 for the combined data (U, , Y, ) may be similarly defined. A proof of 9 2 9, + 9, (2 in the sense of nonnegative definiteness) is given, and equality is interpreted. :J = 9, + 9, if and only if all best linear unbiased estimators (BLUE's) from the co…

expand abstract