Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: separate multiple imputations within the same dataset


From   "Neal Traven" <[email protected]>
To   "Statalist" <[email protected]>
Subject   st: separate multiple imputations within the same dataset
Date   Wed, 14 Jul 2010 14:15:30 -0700

Hello StataList,

I have a fairly large dataset (about 10K records) containing healthcare data from both pre-hospital and hospital settings. Since I don't have Stata11, I'm using -ice- for multiple imputation.

My problem is that I want do two separate imputation runs in the same dataset -- one that uses the pre-hospital variables to impute missing pre-hospital data, and one that uses hospital variables to impute missing hospital data. Each of those imputation runs could also use patient data (age, gender, etc.). Though I haven't thought it through yet, I might also want to use some of the pre-hospital variables as predictors in the hospital imputations; for example, whether a patient was intubated on the way to the hospital might affect some of the hospital variables.

If I use the -m()- option to do, say, 5 imputations on the pre-hospital variables, how can I use those results for the hospital variables? I don't want to make a dataset with 36 times the original size of the dataset (that's the original plus 5 imputed datasets, squared).

What I would really like to be able to do is to run both imputations in a single Stata command, telling it to use the pre-hospital variables for pre-hospital imputation and the hospital variables for hospital imputation ... at the same time.

Is that possible? Is there some way to do this? Am I making any sense at all?

----------
Neal D. Traven, PhD
Research Consultant
Harborview Injury Prevention & Research Center
325 Ninth Ave, Box 359960, Seattle WA  98104-2499
206-744-9459, www.hiprc.org

CONFIDENTIALITY NOTICE: This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential and privileged information protected by law. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies of the original message. To view our complete Notice of Privacy Practices, visit our web site at www.uwmedicine.org.
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index