Accros model hypothesis.

[Henk Hollander]

Hi everyone,

So I have 2 groups in my panel data regression which I would like to say that an independent variable is signifcantly different between them.
So:
model 1: xtreg Y X1 X2 if group ==0, fe
estimate store m1
model 2: xtreg Y X1 X2 if group ==1, fe
estimate store m2
H0: No difference between the X1s
HA: Difference between the X1's

(Note that the variables are the same in the models)

Then I wish to say that model1 X1 != model2 X1 and back it up by a statistical significance test.


My second wish is to have one table, with both the models (m1 and m2) and the difference test between them (m3).
How to do so?

I was thinking along the lines of esttab m*
But have been unable to figure it out.

Thank you in advance
Yours sincerely

[Carlo Lazzaro]

Henk:
see https://www.statalist.org/forums/for...y-different-ys (your first wish).

[George Ford]

suest or just include all the data and interact the group dummy with the variable you want to test for a difference.

reg y x1 x2 group c.group#c.x2

[George Ford]

suest or just include all the data and interact the group dummy with the variable you want to test for a difference.

reg y x1 x2 group c.group#c.x2

[George Ford]

suest or just include all the data and interact the group dummy with the variable you want to test for a difference.

reg y x1 x2 group c.group#c.x2

[George Ford]

suest or just include all the data and interact the group dummy with the variable you want to test for a difference.

reg y x1 x2 group c.group#c.x2

[George Ford]

Carlo pointed you to suest. or just include all the data and interact the group dummy with the variable you want to test for a difference.

reg y x1 x2 group c.group#c.x2

[George Ford]

Carlo pointed you to suest. or just include all the data and interact the group dummy with the variable you want to test for a difference.

reg y x1 x2 group c.group#c.x2

[George Ford]

suest or just include all the data and interact the group dummy with the variable you want to test for a difference.

reg y x1 x2 group c.group#c.x2