Nominal variable

A variable is nominal when its categories are names and nothing else, so no arrangement of them is more correct than any other. The test is not what the values look like. It is whether swapping two categories around in a list loses information. Blood type, home state, and brand of phone all pass that test: shuffle the categories and the data still say what they said.

Two hundred blood donors type as O 92, A 76, B 22, AB 10. The summaries you may report are counts, proportions, and the mode: O is the most common type at $92 / 200 = 0.46$, and the four proportions add to exactly 1. There is no median blood type and no mean blood type, because both would need an ordering this variable does not have.

"The values are numbers, so the variable is quantitative." Five players wear jersey numbers 3, 12, 21, 34 and 55. Their mean is $125 / 5 = 25$, a number that identifies nobody and that would change if the team reissued numbers without a single roster change. Zip codes, area codes, student ID numbers and the codes a survey assigns to answer choices are nominal for the same reason: the digits are labels. The general test is in categorical vs quantitative variables.

The reverse trap is just as common, and words do not make a variable nominal. Poor, fair, good and excellent are words carrying a real order, which makes them ordinal and puts a median category back on the table. Nominal is the case where the order genuinely is absent. Because the category order is arbitrary, the bars of a bar graph of a nominal variable can be sorted by height with nothing lost.

The Fall 2026 AP Statistics framework never uses the words nominal or ordinal. It defines one type, the categorical variable, also called qualitative, as one that takes on category names or group labels. Treat the split as useful vocabulary rather than tested content.