Help using Base SAS procedures

Proc MI and monotone regression

Reply
N/A
Posts: 1

Proc MI and monotone regression

Hello,

I am using proc mi with the monotone regression statement to create 20 imputations:

proc mi data=data seed 100090 out=miout nimpute;

var var1 var2 var3 var4 var5 var6;

monotone regression (depvar=var1 var2 var3 var4 var5 var6/ details);

run;

Because I have the details option SAS provides me with this output to 20 for the 20 imputations:

                               Regression Models for Monotone Method

Imputed ------------------------Imputation-----------------------

Variable Effect      Obs-Data            1              2              3             

depvar Intercept    -0.05954      -0.053980      -0.033559      -0.061138     

depvar var1          0.17215       0.195715       0.137263       0.180499      

depvar var2          0.01896       0.026219       0.024122       0.006856      

depvar var3          0.10666       0.104187       0.103720       0.099717      

depvar var4          0.01891       0.031474       0.037398       0.053791      

depvar var5          0.04823       0.030056       0.061551       0.081066      

depvar var6          0.04659       0.051460       0.050588       0.076492      


However, I am not too sure what the actual numbers mean.  I thought they may be standardized coefficients and the obs-data column in close to the observed standardized coefficients, but they are still slightly different.  Any insight on how to get SAS to provide me with the coefficients and standard errors from the regression imputation equation would be greatly appreciated.

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