|
I would also be very interested to learn more details regarding the instances where this approach is or is not acceptable. This approach is nicely explained and demonstrated by Paul Allison in his SAGE text on MI. I have used this approach routinely when an interaction term is important to the testable hypothesis in a multivariable model (i.e., split on the dichotomous X1 variable with little or no missing data, perform MI stratified on the X1 variable and X1 falls out of the MI model, then recombine the stratified MI data for analysis). Thanks.
Craig
Craig D. Newgard, MD, MPH
Assistant Professor Department of Emergency Medicine Department of Public Health & Preventative Medicine Oregon Health & Science University 3181 Sam Jackson Park Road Mail Code CR-114 Portland, OR 97239-3098 (503) 494-1668 (Office) (503) 494-4640 (Fax) [EMAIL PROTECTED] >>> "Patrick S. Malone" <[EMAIL PROTECTED]> 6/9/2004 6:55:40 AM >>> On Wed, 9 Jun 2004 08:04:40 -0400 (Eastern Daylight Time), Rod Little <[EMAIL PROTECTED]> wrote: > Dear Paul: this is an interesting issue. For the specific case you > outline, the X1Y interaction should be included; a simple strategy would > be to simply stratify on X1 and impute Y and X2 separately in the two > strata. That strategy only applies in limited situations though. Rod > Rod, Could you elaborate, please, on the limited situations? I use that approach fairly routinely in intervention studies -- imputing separately by condition, to preserve treatment X initial status (or whatever) interactions. Thanks, Pat Malone -- Patrick S. Malone, Ph.D., Research Scholar Duke University Center for Child and Family Policy Durham, North Carolina, USA e-mail: [EMAIL PROTECTED] http://www.duke.edu/~malone/ _______________________________________________ Impute mailing list [EMAIL PROTECTED] http://lists.utsouthwestern.edu/mailman/listinfo/impute |
_______________________________________________ Impute mailing list [EMAIL PROTECTED] http://lists.utsouthwestern.edu/mailman/listinfo/impute
