[Gretl-users] Re: On the use of Fixed Effect and Random Effects models for Panel Data