When comparing two groups, the smaller the variation in two groups, then we will be more confident that whatever conclusions we are drawing will leads to less significant differences in the group. In case if the variations in the groups are less, then the group will be less heterogeneous and hence the conclusions can be drawn with much more confidence
