How to conduct the one-sided Wald test in Stata

[Jaden Chen]

Hi everyone,

I hope you're all doing well. I have a question regarding Stata that I would like to seek your assistance with.

I'm wondering how to conduct a one-sided Wald test on the estimated coefficients _b[State1:World] and _b[State2:World] from a Markov-switching dynamic regression model.

Here is the regression model code:

mswitch dr USA_All , states(2) switch(World) varswitch

And here are the regression results:

The Wald test results for both sides are as follows:

 test _b[State1:World] = _b[State2:World]

( 1) [State1]World - [State2]World = 0

chi2( 1) = 58.91
Prob > chi2 = 0.0000


I need to perform a one-sided Wald test with the null hypothesis:

H0: _b[State1:World] - _b[State2:World] > 0

I guess I need to use the invnchi2tail(df,np,p) command, but I'm not sure how to implement it in this specific context. Any guidance or suggestions would be greatly appreciated!

Thank you very much for your help in advance!

P.S. Attached are the data and relevant code.

Best regards,

Jaden

[Dimitriy V. Masterov]

Take a look at the last example at https://www.stata.com/support/faqs/s...-coefficients/

[Jaden Chen]

Take a look at the last example at https://www.stata.com/support/faqs/s...-coefficients/

Thank you so much for your reply！