Programming the statistical procedures from SAS

MIANALYZE and Stepwise Logistic Regression

Reply
N/A
Posts: 0

MIANALYZE and Stepwise Logistic Regression

I have used PROC MI to impute missing data in a dataset (10 imputations). I then run the following syntax:

proc logistic data=impute.ANALYSIS20_4 desc ;
model peer_sex_perp = sex grade1 grade2 race1 race2 race3 famstatd
peer_sex_vict d_delinq gang hed1 hed2 aoi1 aoi2 drug d_efficacy
d_depress d_peerviol d_dateviol d_parpos d_parmon d_peerdel
d_socsupp d_commviol witness cm
/expb covb selection = stepswise include = 8 ;
by _imputation_;

ods output ParameterEstimates=pv_out.pv_parms_multi
CovB=pv_out.pv_covb_multi;
run;

When I go to run PROC MIANALYZE, though, I tried listing all the variables in the model statement above on the modeleffects line, but that didn't work. I'm not sure if this is a really simple question, but I have not been able to find suggestions / advice on how to proceed.

What DO I put on the modeleffects line in MIANALYZE when I use stepwise logistic regression?

Thank you for the help.

Merle
Ask a Question
Discussion stats
  • 0 replies
  • 83 views
  • 0 likes
  • 1 in conversation