A marketing demographic survey is undertaken to determine the market potential for a new product. One of the questions asked is: What type of residence do you live in? Respondents are offered several possible answers including: house, apartment, or condominiums. In order to computerize the survey answers, the responses are coded as a 1 if the answer is "house", a 2 if the answer is an "apartment", and a 3 if the answer is a "condominium". These numbers, 1, 2, and 3, are examples of which level of data?

The numbers 1, 2 and 3 are examples of the level of measurement called as nominal. In this case the numbers are just labels or categories and getting an average value for example does not make sense. Besides, we can't say in general that a house is better than an aparment for example. In others words, we can't find an order between these "labels", and because we are dealing only with "names" we can't compute a variance or mean, but we can find a mode or maybe compute frequencies or percentages.