Analysis of Variance (ViSta-ANOVA) is a technique for comparing means of several populations. ViSta can perform one-way, two-way, or n-way analysis of variance with optional two-way interactions. The samples from the populations may be all the same size (balanced) or may be different sizes (unbalanced). There must be at least one observation in each cell (the data must be complete). 

In one-way ANOVA samples are drawn from each population and the data are used to test the null hypothesis that the population means are equal. 

In two-way ANOVA the populations ar_e classified in two ways. In n-way ANOVA the populations are classified in three or more ways.  For these types of ANOVA, the analysis still compares the means of populations. However, in two-way and n-way ANOVA several comparisons of the sample means are possible since there are several classifications. In particular, comparisons are made for the classifications themselves (called "main effects") and for the interactions of each pair of classifications (called "two-way interaction effects"). ViSta-ANOVA does not compare interaction higher than two-way.

ViSta-ANOVA provides significance tests to test the null hypothesis that the group means, as classified, are all equal. For the significance tests to be accurate we must assume that the samples are drawn independently from normally distributed populations which have the same standard deviations.

The ViSta-ANOVA visualization presents plots to help you assess the assumptions and to help you understand the results of the significance tests. 
