From this discussion, one way of conducting the subsample test is changing the equation from

Dependent_variables= pt + Independent_variables + fixed effects + error term


Dependent_variables= pt + developed_dummy * pt + Independent_variables + fixed effects + error term

while developed_dummy equal to 1 if the firm in developed countries and 0 otherwise I am wondering whether it is the case?

And if it is the case, how come we explain the meaning of the coefficient of developed_dummy*pt ?


I'm a little unsure of what "pt" means, but I assume that "pt" is the independent variable of interest.

In the regression with the interaction term, the coefficient for "pt" is "the expected value of the dependent variable for undeveloped countries if all other independent variables are 0".

The coefficient for the interaction term is "the average difference in the outcome variable for developed countries compared to undeveloped countries, holding fixed all other independent variables".


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.