I run the same regression for two groups (group A vs group B).That is,
Y = a1 + b1X (1)
Y = a1 + b1X (2)

where in equation (1), the regression is run in group A and in equation
(2), the same regression is run in group B.
I would like to test whether b1 in equation (1) is larger than b1 in
equation (2). I know that I can use an interaction term b1*group A dummy
to combine both groups. But if I cannot combine both groups, is there any
formal test to test whether b1 in (1) is larger than b1 in (2)?

I appreciate any help that you may provide.

On Mar 9, 3:53=A0pm, lucy.c...@ASU.EDU (Lucy Chen) wrote:
> I run the same regression for two groups (group A vs group B).That is,
> Y =3D a1 + b1X (1)
> Y =3D a1 + b1X (2)
>
> where in equation (1), the regression is run in group A and in equation
> (2), the same regression is run in group B.
> I would like to test whether b1 in equation (1) is larger than b1 in
> equation (2). I know that I can use an interaction term b1*group A dummy
> to combine both groups. But if I cannot combine both groups, is there any
> formal test to test whether b1 in (1) is larger than b1 in (2)?
>
> I appreciate any help that you may provide.

Why can't you create a dummy variable?