Chow test for a specific coefficient?

Bhav

New Member
#1
Hi there,

I've run an OLS regression on two samples: 1998 and 2004.

My primary interest is the coefficient on the male dummy - I'm trying to see if the change in that coefficient between those two years is significant.

I'm aware that a Chow test can test if all the coefficients are significantly different. But what about testing if just the coefficient on the male dummy has changed?

I believe I can use slope dummies, as shown here: http://kb.iu.edu/data/chow.html

Is this the correct/best way of going about it?

In my case, I'll be creating a dummy that's 1 or 0 (for 1998/2004) and multiplying it with another dummy that's 1 or 0 (male/female). Does that cause any problems for the approach above?

Thanks

PS: I have tried what that website says, and I've got a result. I'm just unsure if it's correct or the most appropriate method in my case.

Edit: forgot to mention that I'm using Stata.

Edit2: nevermind, figured it out using http://www.stata.com/support/faqs/stat/chow.html
 
Last edited: