Beta regression problem

[Muhammad Abid Shahzad]

Hi, I am using panel data for 27 different countries in 5 different time periods. My IV and DV both are in fractional form (0,1), i.e., percentages.
Please guide me, if beta regression is appropriate or I should go with a fixed-effect model, as the T is 5?
If I use beta regression, how can I get country-fixed effects?

Thanks for the help in advance.

[Joro Kolev]

Linear fixed effects model is appropriate in this situation so this is where you should start.

For nonlinear variations, look at this thread here: https://www.statalist.org/forums/for...-fixed-effects

Beta regression is appropriate in principle for your case, but I do not know how fixed effects work with a beta regression, if they work at all.

[Muhammad Abid Shahzad]

Linear fixed effects model is appropriate in this situation so this is where you should start.

For nonlinear variations, look at this thread here: https://www.statalist.org/forums/for...-fixed-effects

Beta regression is appropriate in principle for your case, but I do not know how fixed effects work with a beta regression, if they work at all.

I am sorry, I am not very technical at the concepts of economics. So, by linear fixed effects, you mean xtreg, fe? right?

[Joro Kolev]

Yes, linear fixed effects model with country fixed effects would be either

xtset country
xtreg y x, fe

or directly

areg y x, absorb(country)

I am sorry, I am not very technical at the concepts of economics. So, by linear fixed effects, you mean xtreg, fe? right?