Homogeneity vs Homoscedacity

Greek roots help us better understand hard to pronounce statistical terms such as Homogeneity and Homoscedacity. Simply stated "homo" means same and "scedacity or skedastikós" means to scatter and "geneity" means kind or stock. Homoscedacity compares the means of multivariate (two or more groups) for equal levels of variability, while homogeneity is used for the same value of the mean of univariate analyses.

Homogeneity of Variance

A requirement for the ANOVA test is that the variances of each comparison group are equal and you use the Levene’s Test to determine this. What you’re looking for here is a significance value that is greater than .05. You don’t want a significant result, since a significant result would suggest a real difference between variances. For example, the significance value of the Levene statistic is .155. This is not a significant result, which means the requirement of homogeneity of variance has been met, and the ANOVA test can be considered to be robust.

contributed by Sheri Prendergast