topic Chi-square for categorical variable with more than 2 responses in Statistical Procedures

Chi-square for categorical variable with more than 2 responses

ssulli11 — Tue, 11 Dec 2018 02:52:51 GMT

I am trying to perform a chi-square test on a categorical variable with more that two responses.

For example Variable Age has three responses:

0: 18-29

1: 30-39

2: 40-49

I just want to perform the Chi-square for only the 30-39 age group between my two cohorts. I am sure there is a simple code for this, but I as of yet have been unable to find it.

Thank you for all your help!

Re: Chi-square for categorical variable with more than 2 responses

PGStats — Tue, 11 Dec 2018 04:55:45 GMT

You can restrict testing to a subset of your data with a WHERE statement. Example:

proc freq data=myData;

where age = 1;

.... some test ...

run;

Re: Chi-square for categorical variable with more than 2 responses

ssulli11 — Sun, 16 Dec 2018 18:22:43 GMT

Thank you for your response!

Yes, this is how I originally thought to complete this as well. However, the results are only giving me the table, not the chi-quare results with this code.

proc freq data = have;
where age2 = 1;
tables overlap2*age2 / chisq nocol nopercent;
run;

Re: Chi-square for categorical variable with more than 2 responses

Rick_SAS — Mon, 17 Dec 2018 11:14:08 GMT

Remove the age2 variable from the TABLES statement:

proc freq data = have;
where age2 = 1;
tables overlap2 / chisq nocol nopercent;
run;