Accession Number : AD0708165

Title :   ON THE EXTENSION OF GAUSS-MARKOV THEOREM TO SUBSETS OF THE PARAMETER SPACE UNDER COMPLEX MULTIVARIATE LINEAR MODELS,

Corporate Author : COLORADO STATE UNIV FORT COLLINS

Personal Author(s) : Srivastava,J. N. ; McDonald,Lyman

Report Date : MAR 1970

Pagination or Media Count : 23

Abstract : The paper concerns the problem of linear estimation (without the assumption of normality) under certain general kinds of multiresponse linear models. These include the general incomplete multiresponse (GIM) model and its important special case, the hierarchical multiresponse (HM) model, and also the multiple design multiresponse (MDM) model. These were considered in Srivastava (1967), where the general problem of obtaining the best linear unbiased estimate (BLUE) of general linear functions of the location parameters was investigated. The study is continued here in the direction of obtaining necessary and sufficient conditions for each of the above models to permit the existence of BLU estimates for all elements in a subset of the set of all estimate linear functions of the location parameters. (Author)

Descriptors :   (*MULTIVARIATE ANALYSIS, DECISION THEORY), SET THEORY, EXPERIMENTAL DESIGN, THEOREMS

Subject Categories : Statistics and Probability

Distribution Statement : APPROVED FOR PUBLIC RELEASE