Am 26.12.2021 um 09:33 schrieb Artur T.:
Am 26.12.21 um 08:52 schrieb 3J LEMA:
> Hi, *Gretl and Panel Data Analysis Experts.*
> Using Gretl, how to perform Fixed-Effect and Random-Effects models for
> panel data with a Dichotomous DV?
Hi Lema,
what does "Dichotomous DV" refer to? Could you elaborate a bit, please?
DV must be "dependent variable", so it's a binary dependent. So we're
talking about panel logit or probit, basically.
For random-effects probit go to Model/Limited dependent/Probit/Random
effects. In a script, add the --random-effects option to the "probit"
(not the "panel") command. There's also an example script reprobit.inp
shipped with gretl.
For fixed-effects logit there's Jack's function package "felogit". If
that's installed and attached to the menus, it s under Model/Limited
dependent/Logit/Fixed-effects logit. See the package's documentation for
how to use it in a script.
AFAIK, gretl doesn't offer RE-logit, because it's rarely used. (And just
as a reminder, FE-probit is inconsistent, so it's not given, either.)
cheers
sven