Misuse of multiple comparison tests in some animal nutrition experiments

Abstract

Object of some animal nutrition experiments related with graded levels of the orthogonal quantitative factor is to determine the mathematical relationship between orthogonal and non-orthogonal factors. Researchers are interested in effects of different levels of a quantitative factor on dependent variable and optimum levels of that factor. Quantitative factor levels are depended on each other and continuous. Multiple comparison test can be controlled only in determined levels. In the experiments where independent quantitative factor levels are main target, instead of concentrating on independent quantitative factor levels, making use of relationship between is more reasonable. Regression analysis should be applied instead of multiple comparison tests misused for such researches.