"More Coefficients than Level Specified" error

New Contributor
Posts: 2

"More Coefficients than Level Specified" error

I'm conducting a mediation analysis and am running the total effect of the model.

It is a sample of ~3000 high school students from 10 different schools. I'm getting the error, bolded below, and I can't seem to figure out why. 


The IV (Par_Edu_R) is highest parental education and the DV is a binary substance use variable. There are 6 levels of the IV and it is ordinal (1-6).


I still get parameter estimates, but I want to make sure it's being estimated correctly given the errors. Any help anyone can provide would be much appreciated Smiley Happy


 title 'DV = lifetime use, M = sub reinf, sample = whole thing; IV --> DV- Total Effect among Whole Sample';

 proc genmod data = srnt_alt_r descending;

 where W2_PES_SR_Product_wSum ne . and W3_Life_Use_Any_Sub ne . and  W1_DEM_High_Par_Edu_R ne .;

CLASS W1_School;

MODEL W3_Life_Use_Any_Sub = Par_Edu_R / link=logit DIST=bin ;

repeated Subject=W1_School / type=cs;

estimate 'odds ratio for exp' Par_Edu_R 1 -1/ exp;



AND I'm getting this error:



NOTE: PROC GENMOD is modeling the probability that W3_Life_Use_Any_Sub='1'.

NOTE: Algorithm converged.

NOTE: Algorithm converged.

NOTE: The empirical covariance matrix estimate is used in the ESTIMATE statement.

WARNING: More coefficients than levels specified for effect par_edu_R.  Some coefficients will

         be ignored.

NOTE: PROCEDURE GENMOD used (Total process time):

      real time           0.37 seconds

      cpu time            0.31 seconds

Ask a Question
Discussion stats
  • 0 replies
  • 1 in conversation