You are not currently logged in.
Access JSTOR through your library or other institution:
Built-In Restrictions on Best Linear Unbiased Predictors (BLUP) of Random Effects in Mixed Models
Shayle R. Searle
The American Statistician
Vol. 51, No. 1 (Feb., 1997), pp. 19-21
Stable URL: http://www.jstor.org/stable/2684686
Page Count: 3
Preview not available
In the usual mixed model of analysis of variance we show that certain sums of best linear unbiased predictors (BLUP) of random effects are zero. Those sums are similar to, but not exactly the same as, those of the Σ-restrictions sometimes used for fixed effects.
The American Statistician © 1997 American Statistical Association