- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
I am trying to use proc surveyreg to run a t-test for my complex survey data, but I keep getting an error stating " Variable Depression_Indicator in list does not match type prescribed for this list/ No Model Statement"
This is the code I am using
PROC SURVEYREG data=final nomcar;
STRATA sdmvstra;
CLUSTER sdmvpsu;
CLASS riagendr;
WEIGHT MEC4YR;
MODEL DEPRESSION_INDICATOR = riagendr/solution vadjust=none;
run;
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Is DEPRESSION_INDICATOR a character variable? What does PROC CONTENTS say?
To do a t-test, your Y variable must be numeric.
Paige Miller
- Mark as New
- Bookmark
- Subscribe
- Mute
- RSS Feed
- Permalink
- Report Inappropriate Content
Given the name of your variable I am assuming that it is a categorical variable. To compare groups for a categorical response with complex survey data you should use Proc SURVEYFREQ instead with one of the Chi-Square tests (e.g. CHISQ option on the TABLES statement.
PROC SURVEYFREQ data=final nomcar;
STRATA sdmvstra;
CLUSTER sdmvpsu;
WEIGHT MEC4YR;
TABLES DEPRESSION_INDICATOR*riagendr/CHISQ;
run;