Re: test: changing significance levels

[Richard Williams <Richard.A.Williams.5@nd.edu>]

At 02:24 PM 3/24/2005 -0500, Yvonne Capstick wrote:
> I apologize for such a simple question. I am using the "test" command to 
> test for equality of coefficients under an F-test, having just performed a 
> regression. Is there any way of either:
>
> a) changing the command to test that one coefficient is greater than the 
> other (i.e. specifiy a one-sided alternative hypothesis)
> b) specifying a different signficance level than 5%. Unlike for other 
> commands, an option ", sig" does not seem to be available for "test".
>
> I've tried "help test" but it hasn't listed anything that can help.
1.  Look at the coefficients and see if the difference is in the 
hypothesized direction.  If so, then

2.  Take the reported probability and divide by 2.  e.g. if it says Prob > 
F = .06, the one-tailed probability is .03.

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/