Hello Everyone,

 

I’ve been doing a little reading about Monotone Data Augmentation using the Monotone Data MCMC Method in SAS Proc MI. I understand that the procedure involves imputing enough just enough values for certain variables to produce a monotone pattern of missingness and then using a regression method for monotone missing data to impute the remaining missing values. My sense is that there are also other methods of implementing this two step process.

 

According to the SAS Online Documentation for PROC MI, the Monotone Data MCMC Method:

 

"… is useful especially when a data set is close to having a monotone missing pattern. In this case, the method only needs to impute a few missing values to the data set to have a monotone missing pattern in the imputed data set. Compared to a full data imputation that imputes all missing values, the monotone data MCMC method imputes fewer missing values in each iteration and achieves approximate stationarity in fewer iterations."

 

So it seems clear that the approach works well when the data approximate a monotone pattern. I have reasons for using the approach that go beyond the normal consideration of whether the data approximate the monotone pattern though. So I was wondering if there is any reason to believe that a Monotone Data Augmentation approach will work poorly when the data deviate substantially from the pattern. And if so, what would constitute a substantial deviation?

 

Thanks,

 

Paul

Paul J. Miller, Ph.D.
Research Scientist and Statistician
Ontario HIV Treatment Network
1300 Yonge St., Suite 308
Toronto, Ontario M4T 1X3
Phone: (416) 642-6486 ext 232
Fax: (416) 640-4245

 

_______________________________________________
Impute mailing list
[email protected]
http://lists.utsouthwestern.edu/mailman/listinfo/impute

Reply via email to