[Gretl-users] Re: Fwd: FIML estimation in single equation regression
Am 30.04.2020 um 18:38 schrieb José Perles:
Thanks Sven for your fast response. In fact I have a sample with
lots of missing data. From 207 observations I get with OLS command an
effective sample size of 76. I have seen that FIML would be a
solution and some perform the analysis for example in Stata using
“sem” command (method=mlmv). I would know of it is possible to
replicate the Stata result in Gretl. Thanks again José
OK I briefly looked at what sem/mlmv is supposed to do in Stata, and
first of all I like that they don't call it FIML. Because as you saw
FIML is also used in other contexts such as multiple equations
estimation. So this is a specific way of handling data that is "missing
at random".
I'm not aware that this estimator is available in gretl or in a
contributed package, sorry.
Of course, being an ML estimator one could use the ML (mle ... end mle)
interface that exists in gretl, see also ch. 24 of the guide. There you
would have to write down the likelihood, but I'm guessing you were
looking for an easier off-the-shelf way.
cheers
sven