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 on April 23, and its replacement, is already up and running.

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

RE: st: mi impute chained error messages

From   chong shiauyun <>
To   <>
Subject   RE: st: mi impute chained error messages
Date   Wed, 24 Oct 2012 09:12:21 +0800

I used the options 'chainonly savetrace(impstats) dryrun report' 
in my MI model but no iterations of performed, though the result screen 
was showing 'performing chained iterations'. 

Can someone please advice? Thanks


> Date: Tue, 23 Oct 2012 12:58:45 -0400
> Subject: Re: st: mi impute chained error messages
> From:
> To:
> On Tue, Oct 23, 2012 at 2:26 AM, chong shiauyun <> wrote:
> > Thanks for your response. I will try linear regression or I may use truncated regression that allows me to specify the lower and the upper limit. Actually I did manage to run the dryrun when I put my syntax like this
> > mi impute chained (reg) .... (ologit)... (intreg, ll(lTIQ) ul(uTIQ))TIQ, add(20),... dryrun report (just have to put the lower and upper limit in the same blanket as the (intreg) and then specify a new variable, TIQ, that have the impputed value of IQ.
> >
> > I think truncated regression would be more suitable to apply to my IQ scores (which are continous variables with boundaries between 45-155).I wil also try linear regression.
> I suspect you won't have too many near the boundary so don't sweat it.
> It makes the imputation more complicated.
> Also, don't overlook predictive mean matching (pmm) as a method when
> you have boundaries and such, as long as the N is large.
> > Another thing that I would like to know is that does it matter if I put variables that I have collapsed (eg. for education level, I collapsed bachelor degree and postgraduate degree into a group called high education) instead of the original data into the imputation model?
> Ideally the imputation model mirrors the behavior in the data as well
> as possible. This means that it's often more complex than the intended
> analysis model, to some degree, and may include things like nonlinear
> terms for a given variable when you plan on using a linear term in the
> model, or extra interactions. Dirty cases or variables have their
> influence spread all over the dataset, though, just like in any other
> full information method, so be wary.
> *
> *   For searches and help try:
> *
> *
> *
*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index