Experimental, quasi-experimental and ex post facto designs often generate results that are not interval data, continuous numbers, but nominal or ordinal data instead. Thus parametric tests are not the appropriate tools to resolve the hypotheses. There is a need for tests ...