On Additivity in the Analysis of Variance

Publisher Website
Google Scholar
Abstract
The problems of testing for non-additivity and of finding the best transformation of the data to minimize it are discussed. Consider the model: E(yq) = [mu] + [mu]q + c[mu]2/q. , where yq is an observation belonging to the q-th subclass in an analysis of variance problem. S this model fits our data, and if max uq - min [mu]q < 1/

This publication has 0 references indexed in Scilit: