Re: st: Goodness of fit for discrete hazard model with unobserved heterogeneity

[Maarten Buis <maartenlbuis@gmail.com>]

On Sun, Jul 3, 2011 at 6:35 PM, Urmi Bhattacharya wrote:
> I am estimating a discrete hazard model with unobserved heterogeneity
> ( which is statistically significant). Is there any test for goodness
> of fit for such a model? I generated the fitted survival function and
> compared it with the sample survivor function but it is hardly
> satisfactory because a) using the below model I get a predicted hazard
>  which assumes that the random effect is 0 and

If you use -xtmelogit- instead of -xtlogit- you can estimate exactly
the same model, but afterwards you can also predict the random
effects. Together with the predicted linear predictor you can turn
these into individual level probabilities.

> b) It is only a visual comparison and hence not very convincing.

I could not disagree more. When it comes to checking models test are
the most unconvincing pieces of evidence and graphs the most
convincing. Think about what a model is: it is a simplification of
reality. This is just another way of saying that models are wrong,
useful but wrong. So we do not need to do any tests as we already know
the answer: We should reject the null hypothesis and if we don't our
sample is just too small. The real issue with choosing models is to
find the balance between simplicity and realism. This is a judgement
call, where graphs are excellent tools to help you make those
decisions and document them so you can explain your reasons to others.

Hope this helps,
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/