Date:         Thu, 17 Feb 2011 09:33:15 -0500
Reply-To:     William Shakespeare <shakespeare_1040@HOTMAIL.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         William Shakespeare <shakespeare_1040@HOTMAIL.COM>
Subject:      Re: Imputing categorical variables

Handling continuous data is straightforward.  The imputation of
categorical data is equally clear.  What's not clear is how to handle the
imputed categorical data.  If one leaves the categorical data continuous
after imputation, then there must be some way to handle it in subsequent
analyses.  Nothing that I read is very clear on this, but what of what
I've seen, the implication is to use a covariance or correlation/mean
matrix.  This makes sense to me but I looked through the SAS docs and I
didn't see anywhere how to use such a data set in, say, proc glm.  Doesn't
mean it's not there.  I just didn't see it if it is.