In an uncontrolled t test, the mediator variable in the experiment has the chance to negatively impact the correlation between the independent and dependent variables. This can cause false relationships to be formed which can lead to an improper analysis of results. This will lead to incorrect rejections of a given null hypothesis.

On the other hand, a multiple comparison test, such as Tukeys method, is a controlled experiment. The conclusions in these tests will lead to an accurate analysis of results and will give us proper rejections of a given null hypothesis. Multiple comparison tests are important in the design of experiments because they can adequately tells us exactly which groups significantly differ from the others.