Re: st: Re: A fractional logit can't converge. Anyway, backed up results are useful for mfx? how?

[Roger Cremades Rodeja <ecosafor@gmail.com>]

Dear Marteen and Nick,

I am extremely grateful for your help and clarity. Using
-technique(bhhh)- it worked perfectly.

Marteen, yes I do have a big spike, you were right! I use proportion
data of 1,300 households, but most of them have a 0 in my dep var, so
I though it was because of that, as in my other two models just change
the dep var and it has more variation.

Nick, I meant logit (I spelt tobit wrongly).

Bless you!!
Roger


El día 8 de marzo de 2012 14:29, Maarten Buis <maartenlbuis@gmail.com> escribió:
> On Thu, Mar 8, 2012 at 1:14 PM, Roger Cremades Rodeja
> <ecosafor@gmail.com> wrote:
>>  I am using fractional tobit "glm depvar indepvar, link(logit)
>>   family(binomial)  vce(robust)". I successfully used it in other two models,
>>  but in this case after some iterations Stata says "backed up" and cannot
>>  converge, then I create a table with the results (outreg), and still appears
>>  a resulting table with results that are consistent with my other two models,
>>  do those results makes statistical sense?
>
> They make no sense, so don't look at them and certainly don't try to
> publish anything with them.
>
>>  I want to create then elasticities with "mfx" but I can't because the
>>  model did not converge, if those results produced with outreg (previous
>>  line) make a bit of sense, how can I create elasticities with them using
>>  Stata?
>
> Don't do that.
>
>>  Also, after typing the command appear "note: perc_mod_com has noninteger
>>  values", I do not understand why, as I think proportion variables should be
>>  0 to 1.
>
> That message makes sense: remember that -glm- is a general purpose
> program that can be "tricked" into estimating a fractional logit.
>
>> Following previous emails in Statalist, I tried delete some
>>  variables and I still couldn't solve it, OLS do not reject any variable, the
>>  model do not work trying to delete indep variables one by one, and I am
>>  trying more and more combinations, but it still does not work.
>
> I would look for some form of perfect prediction. This becomes more
> likely if you have a very large spike in the dependent variable
> (typically at 0 or 1) or a small sample size.
>
>>  Please somebody knows if this analysis can be carried on using another
>>  sofware? I need these results..
>
> I think it is a problem with your data not your software.
>
> -- Maarten
>
> --------------------------
> Maarten L. Buis
> Institut fuer Soziologie
> Universitaet Tuebingen
> Wilhelmstrasse 36
> 72074 Tuebingen
> Germany
>
>
> http://www.maartenbuis.nl
> --------------------------
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/