Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | "Scharnigg, Stan (Stud. SBE)" <s.scharnigg@student.maastrichtuniversity.nl> |
To | "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu> |
Subject | st: wald test, one-sided test not appropriate |
Date | Tue, 7 Jun 2011 19:48:15 +0200 |
Dear statalisters, I have the following regression: regress var1 var2 var3, where var1, var2, var3 are B1, B2, B3 respectively. Hypothesis: H0: B1 + B1 + B1 >= 0 Someone told me that I could use a Wald test to test this. I used the FAQ: http://www.stata.com/support/faqs/stat/oneside.html then, I did the following: regress var1 var2 var3 local sign_var1 = sign(_b[var1]) local sign_var2 = sign(_b[var2]) local sign_var3 = sign(_b[var3]) display "Ho: coef >= 0 p-value = " 1-ttail(r(df_r),`sign_var sign_var2 sign_var3'*sqrt(r(F))) However, at the end of the page there is written: "Again, this approach (performing a Wald test and using the results to calculate the p-value for a one-sided test) is appropriate only when the Wald F statistic has 1 degree of freedom in the numerator" My degrees of freedom is 3 in this case. So, this test is not appropriate. Why is it not appropriate and how could I solve this? Thank you very much * * 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/