Make frequency tables of text based variables

My participants filled in a questionnaire which had 2 pooling variables ('Sex' and 'Age') and 80 other questions of which some answers were numbers (eg 'PhysicalActivityPerDaY' with 0, 1, 2, 3, 4 as possible answers) and some answers which were text-based (eg 'JobSecurity' with High, Not Sure, Low as possible answers).

I want to present the descriptive statistics of all variables per group, so the outcome of the code for variable 'JobSecurity' should have two tables which look roughly like this:

Sex JSHigh JSNot Sure JSLow

male 70% 15% 15%

female 80% 5% 15%

AgeGroup JSHigh JSNot Sure JSLow

18-35 20% 40% 40%

36-45 30% 30% 40%

46-55 50% 40% 10%

56-65 50% 30% 20%

I've tried experimenting with the varfun and grpstats functions, but they both seem unable to work with text-based variables. Is there a way around this?

Thanks in advance for your help

PS: Unfortunately the data collection has already been done by a previous group, so changing the way in which the data is stored (1,2,3 instead of high, not sure, low) is not an option.

Answer 1

[Cam Salzberger]

Hello Jesse,

This is precisely what the categorical data type is for. Once you have converted the text data in the table to categorical, you can essentially treat them as numeric when taking statistics and such.

Hope this helps!

-Cam