Vectors, a tool in statistical regression theory

摘要: Using linear algebra this thesis developed regression analysis including of variance, covariance analysis, special experimental designs, and fertility adjustments, experiments at different places times. The determination the orthogonal projection, yielding efficient unbiased estimates playing a dominant role in statistical tests, was extensively considered, particular by iterative time-saving procedures based on geometrical considerations power-series expansions. Subspaces were systematically introduced for levels, main effects interactions. This allowed general interpretation orthogonality classifications. introduction subspaces also permitted simultaneous account polynomia one or more variables with concurrent Designs such as balanced partially incomplete block lattices Latin squares could be characterized fixed angles between spaces physically meaningful these effect spaces. missing-plot technique simple form. Finally, estimation testing problems discussed several non-orthogonal classifications, which had random included recovery inter-block information, split- plot designs series experiments.