if our data are on a Nominal or Ordinal Scale they require NON parametric tests
WHY?
Parametric tests require numerical scales where the numbers and intervals have mathematical meaning
Nominal scales (simple labels like for example car's brands) violate the mathematical assumptions of normal distribution or the ability to calculate a mean or a standard deviation required by parametric tests
The same applies to ordinal scales (ordered - ranked categories like 1st 2nd 3rd)
