Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | "Carrion-Menendez, Alejandro" <ACarrionMenendez@imf.org> |
To | "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu> |
Subject | st: xtabond2 with interaction terms |
Date | Tue, 5 Apr 2011 10:20:15 -0400 |
I have a model with continuous interaction terms. I'm having trouble setting up the specification for xtabond2 and I haven't found anything in the literature that refers to this issue. My interactions are basically the product of the constitutive terms (fxratio_ct fxclass1_ct concentrationdummy depgdp_ct) and the variable of interest (policy_ct). Should I follow Roodman's advice of including every regressor into the instrument matrix? Or should I exclude either the interactions or the constitutive terms by default? xtabond2 lending l.lending policy_ct fxratio_ct fxclass1_ct concentrationdummy depgdp_ct fxratio_p fxv_p dep_p cdummy_p year*, gmm(policy_ct lending fxratio_p fxv_p dep_p cdummy_p fxratio_ct depgdp_ct fxclass1_ct concentrationdummy , collapse lag(1 2)) iv( year*)small twostep robust Thanks! * * 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/