ANOVA one-way ANOVA one-way anova is an analysis of variance where the different groups being compared are associated with the factor levels of only one independent variable. The null hypothesis is an absence of difference between the means calculated for each of the groups. The test assumes normality and equivariance of the data.