In the case of nominal data (categories with no implied ordering) we can not determine the average of the categories, and likewise we can not determine what category was the middle, because there is no particular order to the data.

The best we can do is indicate which category was the most frequently recorded. This is called the modal value of the distribution of the data.

For example, if for some reason you were categorizing patients by hair color for the hypertension study, and you found that there were 15 individuals with brown hair, 20 individuals with blond hair, 5 with black hair, 2 with purple hair, and 12 with red hair. This information is plotted on a histogram below.