topic How to combine(pooling) the analysis results after imputation without standard error? Help me in SAS Procedures

How to combine(pooling) the analysis results after imputation without standard error? Help me

LiangLi — Thu, 09 Nov 2023 05:44:47 GMT

I imputed five datasets, and replicated analyse these datasets by clopper-pearson method in SAS Proc Freq.

like this:

proc freq data=imputed;

table aval /binomial(cl=exact);

output out=cl_1 binomial;

by _impute_ arm;

run;

I would get a dataset including the prob(prob_lower,prob_upper) by the clopper-pearson method, but i can not get the standard error.

So i can not continue to finish the Proc Mianalyze in SAS.

Re: How to combine(pooling) the analysis results after imputation without standard error? Help me

Watts — Thu, 09 Nov 2023 15:53:40 GMT

Please see this discussion: PROC MI for Chi-Square

Re: How to combine(pooling) the analysis results after imputation without standard error? Help me

LiangLi — Thu, 09 Nov 2023 21:34:39 GMT

Thanks for the answer, but it can not solve my issue.
In my SAS program, i only have a variable(please view the TABLE statement). And i would calculate the 95% confident intervals by clopper-pearson method. So the CHISQ were not output as it need two variables at least.