Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


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

Re: st: question regarding multiple imputation using ICE command


From   Fred Wolfe <fwolfe@arthritis-research.org>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: question regarding multiple imputation using ICE command
Date   Tue, 9 Mar 2010 07:52:41 -0600

With respect to normality, the help file says,

"With the boot option, steps 2-4 are replaced by a bootstrap
estimation of beta_star and sigma_star, obtained by regressing yvar on
xvars after taking a bootstrap sample of the non-missing observations.
This has the  advantage of robustness since the distribution of beta
is no longer assumed to be multivariate normal."


Fred Wolfe

On Tue, Mar 9, 2010 at 6:07 AM, Katya Mauff <Katya.Mauff@uct.ac.za> wrote:
> Dear all
>
> I am attempting to impute several variables (all MAR as far as I can tell) using the ICE command in Stata, specifically, weight (continuous) and number of mutations (categorical 4 levels).
>
> The data I have on weight is bimodal, and the ICE command requires a normality assumption. I've attempted to determine why the split in the weight data occurs with respect to other available information, and have run a regression of weight on several possible culprit variables (e.g age and pregnancy status). When I run the regression, my residuals are approximately normal.
>
> My first question is thus: if I run the ICE command including all the variables in the earlier regression command, do I still have to normalize weight? (and if so-possible suggestions on how to do this (?) seeing as the split is not determined by any single variable...)
>
> My second question is with regards to the perfect prediction message I get when running the ICE command for the variable indicating mutation numbers- Do I have to use (e.g.) pred_eq or check_eq? Or will the use of augmlogit correct for the perfect prediction in my equation?
>
> Kind Regards
> Katya Mauff
>
> *
> *   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/



-- 
Fred Wolfe
National Data Bank for Rheumatic Diseases
Wichita, Kansas
NDB Office  +1 316 263 2125 Ext 0
Research Office +1 316 686 9195
fwolfe@arthritis-research.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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index