Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Jen Zhen <jenzhen99@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | st: SE of linear prediction: Use only SE of slope coefficient(s), or also that of the constant? |
Date | Wed, 1 Dec 2010 13:55:02 +0100 |
Dear listers, I have estimated Y = A + (B1*X1 + B2*X2 + B3*X3) + (C1*Z1 + C2*Z2+C3*Z3)+eps. The X are a set of mutually exclusive dummies, and the Z are another set of mutually exclusive dummies. Now I want to use these estimated coefficients to compute the average Y for the three different X cases, keeping Z fixed at any one of its values, for simplicity say when all Z are zero. I do this simply by computing A+B1, and similarly for the other constellations of X. But I'm unsure what the correct standard error for this is: Can I simply take the standard error of B1, or do I need also need to add that of the constant A, as well as the covariance between the estimated coefficients A and B1? Many thanks, JZ * * 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/