How do I get a joint P value for 2 separate regressions?

[Damithri Munasinghe]

Hello, I am running a regression to check the difference in health outcomes due to a policy. I want to know the impact this has on gender and I ran two separate regressions for males and females. I can't compare the coefficients because I dont know the joint P value of the two. I want to know if they are statistically different or not. How can I do this?? My regression is reghdfe bmi treat*post treat if female == 0, absorb(REGION year) cluster(CHILDID)

[Carlo Lazzaro]

Damitri:
welcome to this forum.
Why performing two different regressions instead going:

reghdfe bmi treat*post treat i.gender, absorb(REGION year) cluster(CHILDID)

and then using -test- and/or -lincom- for postestimation?

[Damithri Munasinghe]

Hi Carlo,
Thank you for your response!
I have tried the regression sent, and I've attached a screenshot below. My female variable is a binary so is it still ok that I use this command? Also, the command for testing is, test i.female is it?

[Carlo Lazzaro]

Damithri:
1) your code is correct;
2) as per -1.female- coefficient, your regression does not support the evidence that, other things being equal, gender has any bearing on variation of the conditional mena of the regressand -bmi-.
As per FAQ, please do not post screenshots, but use CODE delimiters (that you can access via the #-shaped toggle available from the advanced data editor) to share what you typed and what Stata gave youm back. Thanks.

[Damithri Munasinghe]

Okay, noted. Thanks a lot!