Finally, we consider directly
the issue of “consistency” in the context of the tobit model and find that widely received perceptions to the
contrary, at least in this model, the fixed effects estimator appears to be neither biased nor inconsistent
--- Оригінальне повідомлення ---
Від кого: "Sven Schreiber" <svetosch@gmx.net>
Дата: 6 грудня 2014, 13:23:38
Am 06.12.2014 um 11:38 schrieb oleg_komashko@ukr.net:
> It will be very nice.
> I have a bundle of questions about the future plans, so
> I would be happy being "partially answered"
> tobit marginal effects
don't know
> alternatives to binary choice (quasibinomial, heteroscedastic,
> non/semiparametric)
There is a package on the server by Jack and Claudia for heterokedastic
and IV probit, look for "HIP".
> panel data VAR (as described in Hsiao, for example)
don't know
> panel data probit
There is already the --random-effects option for probit built in. (And
fixed-effects probit doesn't exist, as you may know. But there is
another package by Jack for fixed-effects logit, "felogit". I myself
have recently played around with functions that estimate random-effects
logit by calling R internally and trying to do it with gretl only, but
this isn't really ready for consumption.)
hth,
sven
_______________________________________________
Gretl-users mailing list
Gretl-users@lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users