Description of a categorical variable using textual data

The second method of the Automatic Univariate Description submenu is used to describe categorical variable with textual data.

The interface

Like the other two methods of the Automatic Univariate Description submenu, the GUI is very minimalist and only let you select the variable to describe, the textual variable used for the description and the significance threshold.

Potential issue(s)

Unlike the first Automatic Univariate Description method, you shouldn’t try to set a significance threshold too high. Otherwise, the analysis will run for a very long time (since there might be hundreds of different words) and might not even finish.

Example

The data used to demonstrate this analysis is the beard_description data set (which you can see on the first figure). For this example, we will use the default settings (see the previous figure).