A numeric variable stores one numeric value for each case. Numeric variables are often based on questions that ask "How many?" or "How much?" and require the respondent to give a numeric response.

For example, in the Short Drinks sample data set, respondents are asked to enter their age as a numeric value rather than by selecting the category that represents their age group. When respondents answer the question, they enter a whole number.

Sometimes you may want to group the numeric values stored in a numeric variable into categories. For example, you may want to create a table that shows the age data in age groups. The process of creating categories from a numeric variable is sometimes called banding and each category is called a band. Banding a numeric variable is similar to coding a text variable, but it is generally easier, because you can define the bands using numeric expressions. You can define the bands using UNICOM Intelligence Data Model or you can create automatic bands (for example, ten equal bands starting at the minimum value and ending at the maximum value). See Creating bands for more information.

Numeric variables can be divided into two groups based on the type of numeric data they store: