Kit Baum wrote:
> I'm not sure why transforming variables into logits 'by hand' is any
> more onerous than transforming variables into logarithms 'by hand'
> prior to estimation. And I'm not sure that the specification given --
> something like
>
> glm unemp state public, link(logit)
>
> without the vce() option--makes any sense in a GLM context. -glm- says
>
> no observations
> data not suitable for non-standard family-link combination
> r(459);
>
> The unemployment rate is bounded away from zero, true, but then so
> are 95% of the dependent variables used in econometric analysis. How
> can you fit a logit link using a binomial distribution to this
> relationship?
It can be onerous if you have lots of independent variables to
logit-transform and you have lots of models that you want to fit quickly,
and that's why -glm- would be so handy if one could fit the correct model
with it. But, as you've shown, using -glm- to fit OLS Newey-West models to
panel data is out of the question. So any chance of working such an
automatic option into -ivreg2-, Kit (he asks slyly!)?
CLIVE NICHOLAS |t: 0(044)7903 397793
Politics |e: [email protected]
Newcastle University |http://www.ncl.ac.uk/geps
Whereever you go and whatever you do, just remember this. No matter how
many like you, admire you, love you or adore you, the number of people
turning up to your funeral will be largely determined by local weather
conditions.
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
