Unpaired Analysis of Variance

Abstract

Unpaired t-tests are for assessing two parallel groups of similar age, gender and other characteristics treated differently. However, if you wish to compare three different treatments, three parallel groups are required, and unpaired t-tests can no longer be applied. Instead, unpaired analysis of variance (ANOVA), otherwise called one-way ANOVA must be used to tell us whether there is an overall difference in the data. However, it does not tell us whether treatment 1 is better than 2, 2 better than 3, or 1 better than 3, or any combinations of these effects. For that purpose post hoc tests are required comparing the treatments one by one. Unpaired