Dear John,
Gary King suggests takes log of always positive variables (check his
webpage). Then your logged variable is unbounded and you can impute that
variable instead of the original one. Of course recover the original with
exp().
I hope this helps you
Rodrigo.
----- Original Message -----
From: "John Tabish" <[email protected]>
To: <[email protected]>
Sent: Sunday, June 04, 2006 8:06 PM
Subject: st: multiple imputation
I am trying to do multiple imputation using ICE in stata. My variable to be
imputed is a continuous variable for measuring depression (Hamilton Rating
Scale). When I do the imputation based on some predictor variables and week
number, I am getting negative values for some of the missing values which
are not possible because my variable is always positive. I tried a different
combination of predictors and interaction terms but I keep getting negative
values. Do you have any suggestion or idea to why this is happening?
Any help is appreciated. Thanks
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/