On Analysis of the addition of Center Points to the Unreplicated Full Factorial Design

摘要

One of the most important purposes for experimental design is to analysis the significance of factors. In this paper, the unreplicated full factorial design with the center points will be in progress of the fatorial effects analysis. Considering as the character of the unreplicated experiment , no plenty of degrees of freedom left to estimate the error variance, we use the observed responses at the center point to estimate the mean square error unbiased and analysis the significance of the factors in the normal ANOVA. Otherwise, replicating center points will be provided the test for curvature of the response function over the range of the factor levels chosen. Besides, the paper also introduces the Lenth method, regular test method of fatorial significance of unreplicated experiment, that is given under the condition of no center points.