Numerical and Categorical Data

Shodor > Interactivate > Discussions > Numerical and Categorical Data

Mentor: Without using the precise terminology we have learned about two different types of data. The names for these are "categorical" and "numerical." Can anyone guess what these terms might mean?

Student 1: Well numerical is like number, so maybe that is data with numbers.

Mentor: Very good. If numerical data refers to data that uses numbers, what might categorical data mean?

Student 2: Something to do with categories?

Student 1: Maybe like what you choose from a multiple-choice question. Could that be categorical data?

Mentor: Very good. You are both correct. Categorical data is data that is defined by words or has a limited number of answer options. Let's think of questions that you could ask the entire class and the answers would be considered categorical data?

Student 2: What if we asked everybody about their favorite ice cream flavor?

Mentor: Yes, that would be categorical data because there is a limited number of answer choices.

Student 2: Couldn't there be an infinite number of ice cream flavors?

Mentor: Technically yes, there could be a very large number of ice cream flavors. In this case, we would probably want to condense all of the information into a few categories anyways. Other examples?

Student 1: How about eye color?

Mentor: Very good. We could survey the entire class and find a limited number of eye colors. What are some examples of questions we could ask and get back numerical data?

Student: We could ask for each person's height.

Mentor: That is a great example of numerical data as every person would respond with a number and they could answer with specific numbers , such as 5 feet 4 and one half inches. Any other examples?

Student 1: We could ask everyone for their age.

Mentor: That is a tricky one. If we ask people and they respond "14", "12", etc, that would really be categorical data. The categories would be each age and there would be a limited number of categories. If the class responded with how many days hold they are, we could consider that numerical data as there are a lot more potential answer choices.

Student 1: That is confusing. So there are some questions you could ask and the class would respond with numbers, but it would be considered categorical?

Mentor: I think you've got it! Another example might be if we asked everyone how many pets they have. The answers would all be whole numbers greater than or equal to zero. For data to be numerical the answer choices need to possibly include decimals or fractions.

Student 2: I think I get it!

Mentor: Now you can determine what type of data you have and determine which type of graph is best for your data.

a resource from CSERD, a pathway portal of NSDL NSDL CSERD