Hello everyone!
I am planning to run a fractional regression model in the context of panel data .
As per the literature, to account for unobserved heterogeneity, one must add the time averages of the exogenous variables and estimate the so called "fixed effects fractional regression model". But, how does one estimate a random effects fractional regression model?
Also, in the case of a fractional model, how does one decide between fixed effects and random effects?
Thanks