I sent this email out accidentally a couple days ago to the STATA digest list by mistake. I apologize if you are getting this twice, but I very much appreciate your help.
Hi STATA list,
I am trying to decide whether the use of multiple imputation is an appropriate method for dealing with missing data in my dissertation study, which uses data from the public use Add Health data set. In order to use variables from this dataset, I have had to take into account sample weights, strata, and the primary sampling units as well as specify my subpopulation. I have read some information from a response to a previous listserv question that has lead me to have doubts about whether I should use MI: "-mi- does not provide imputation methods specifically designed to impute complex data, such as panel, longitudinal data, complex survey data, time-series data, etc. The methods employed by -mi- rely on the iid
assumption which is violated in these data, and to the best of my knowledge
the methodologies for imputation methods relaxing this assumption have yet to
be fully developed." I would like to know whether others have used either the new multiple imputation commands in STATA 11 or the user written multiple imputation commands for STATA 9 and 10 with the ADD Health data or with other complex survey data for which you have had to taken into account sample weights, strata, and specify a subpopulation. I would also like to know if others methods for handled missing data when using complex survey data are recommended and which ones. Also, if you have successfully used multiple imputation with similar data, ideas of how long it takes to learn and carryout, and tips on what I need to do for it to work are greatly appreciated.
Thanks!
Karen
*
* 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/