Hi,
I'm using Clarify -estsimp- to estimate a fixed effect model from 5 MI TSCS
datasets created by Amelia 2. I'm curious to know how I can test for serial
correlation like I am able to do with single MI datasets using the -xtserial-
command, but have been unable since -estsimp- does not allow for the -xtserial-
command.
One alternative would be to generate residuals and regress it on the lag of the
residuals (along with other covariates) and check for significance of
correlation. However, the -predict error, resid- command doesn't work after
-estsimp- model output. Please advise.
Thanks,
Brian
--------------------------------
Brian Chin
Research Assistant
Harvard Initiative for Global Health
104 Mt. Auburn St.
Cambridge, MA 02138
U.S.A.
Tel. +1 617 495 4717
Fax. +1 617 495 8231
Email Brian_Chin(a)harvard.edu
www.globalhealth.harvard.edu
--
Clarify mailing list served by Harvard-MIT Data Center
[Un]Subscribe/View Archive:
http://lists.gking.harvard.edu/?info=clarify