Categorical Variables
Index Tutorial Glossary  

A variable is Categorical if it assumes a limited number of cardinal values. A set of all allowed values is the domain of categorical variable. The cardinality property implies that numerical values of categorical variables bear no meaning and serve only as the codes representing the values from the domain. In the Variables box of the Analysis Page, the categorical variables are identified by [C] at the end of the label.

Example

The Common Category Variables (always shown on top in the Variables box of the Analysis Page), such as size, exporter or ownership, are the typical examples of categorical variables.

Notes: In One- or Two-dimension Cross Tabulations, the row and column variables must be categorical, while the analytical variable must be Continuous. To use continuous variables in rows or columns, create the new categorical variables using Group() or BandGroup() functions (see Creating New Variables).

In statistical procedures, such as Means, categorical variables are treated as if each category had a numeric value assigned to it. For example, the range of the variable ownership is {Foreign, Domestic}. The procedure would produce the descriptive statistics as if the category "Foreign" was equal to 1 and "Domestic" was equal to 2.


Related Topics