Two-factor and Multi-factor Analysis of Variance
As explained in the last chapter, single factor ANOVA with blocks is applied to measurements arranged according to two criteria of classification. The measurements relating to the first criterion (Factor A) are replicated, whereas those relating to the second criterion (blocks) are not. The measurements are insufficient to enable one to judge very accurately whether the two criteria act independently or interact. If the variance estimate for error is small, one can safely conclude that any interaction is small. But if the variance estimate is large, one cannot be sure whether there is a large interaction or whether the disturbing influence of other factors is large.
Unable to display preview. Download preview PDF.