st: RE: RE: xtmixed and missing data

["Hoffman, George" <ghoffman@mcw.edu>]

I think you need to reshape your data to long format, with each row representing a unique observation identified by subject and year, and the xtset your data.
Xtreg handles missing values well, but you'll need to be clear about your model assumptions.
I would guess that you then probably want to predict the values at year = 4.

Refer to:
xt data
xtreg


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Dennis Kramer
Sent: Friday, April 29, 2011 5:15 AM
To: statalist@hsphsun2.harvard.edu
Subject: st: RE: xtmixed and missing data

I am estimating an xtmixed function while controlling previous year
experience/achievement. How I am using three previous year testing
data to predict the current year's performance. However, I do not want
to exclude individuals in my sample who do not have a score in each of
three years. Currently, when I am estimating the predicted value, it
is dropping all those who do not have a value in each variable.

Is there is a command I could use to allow data to be estimated on the
number of data points an individual has? Overview of my data is as
follows:

Subject # --- Year X Test --- Year X-1 Test ---- Year X-2 Test ---
Year X-3 Test ---- Covariates


Thank you
*
*   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/

*
*   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/