Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: st: RE: Checking Goodness of Fit after Zero-inflated Poisson


From   Nick Cox <[email protected]>
To   "'[email protected]'" <[email protected]>
Subject   RE: st: RE: Checking Goodness of Fit after Zero-inflated Poisson
Date   Sun, 22 May 2011 19:56:54 +0100

You shouldn't need to ask. Your question was already answered in my post.

As the example output shows, -zip- already shows a LR chi-square statistic and a P-value. 

Nick 
[email protected] 

[email protected]

Thanks Nick for your response.

First, sorry for not understanding the name policy.  I will sincerely
follow the rules.

Ok, after I run a standard Poisson, I can compute the goodness-of-fit
by using the command -estat gof-.  For instance, after I run Poisson,
-estat gof-  reports as follows:

. poisson  day windcat du50 du100 du150

Iteration 0:   log likelihood = -186.58213
Iteration 1:   log likelihood = -183.38329
Iteration 2:   log likelihood = -183.34036
Iteration 3:   log likelihood = -183.34036

Poisson regression                                Number of obs   =        125
                                                  LR chi2(4)      =     606.33
                                                  Prob > chi2     =     0.0000
Log likelihood = -183.34036                       Pseudo R2       =     0.6231

------------------------------------------------------------------------------
         day |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
     windcat |   .6731493   .0625124    10.77   0.000     .5506273    .7956713
        du50 |   2.550244   .3132588     8.14   0.000     1.936268     3.16422
       du100 |   2.144695   .3203497     6.69   0.000     1.516821    2.772568
       du150 |    2.07342   .3292282     6.30   0.000     1.428144    2.718695
       _cons |  -2.008924   .2934786    -6.85   0.000    -2.584132   -1.433716
------------------------------------------------------------------------------

. estat gof

         Goodness-of-fit chi2  =  118.6753
         Prob > chi2(120)      =    0.5170

But, that command does not work after zip.  How may I compute the
goodnes-of-fit? Or, does zip already report the goodness-of-fit in its
result table?

Hope this clarifies my question.

Thanks again,
Sangyub Ryu

On Sun, May 22, 2011 at 2:18 PM, Nick Cox <[email protected]> wrote:
> Dear Spring [in?] Athens:
>
> Please note the request in the Statalist FAQ to use real names.
>
> From a -zip- I see output like this
>
> Zero-inflated Poisson regression                  Number of obs   =        250
>                                                  Nonzero obs     =        108
>                                                  Zero obs        =        142
>
> Inflation model = logit                           LR chi2(2)      =     506.48
> Log likelihood  = -850.7014                       Prob > chi2     =     0.0000
>
> ------------------------------------------------------------------------------
>       count |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
> -------------+----------------------------------------------------------------
> count        |
>     persons |   .8068853   .0453288    17.80   0.000     .7180424    .8957281
>    livebait |   1.757289   .2446082     7.18   0.000     1.277866    2.236713
>       _cons |  -2.178472   .2860289    -7.62   0.000    -2.739078   -1.617865
> -------------+----------------------------------------------------------------
> inflate      |
>       child |   1.602571   .2797719     5.73   0.000     1.054228    2.150913
>      camper |  -1.015698    .365259    -2.78   0.005    -1.731593   -.2998038
>       _cons |  -.4922872   .3114562    -1.58   0.114     -1.10273    .1181558
> ------------------------------------------------------------------------------
>
> What else, precisely, do you have in mind by "goodness of fit"?
>
> Nick
> [email protected]
>
> [email protected]
>
> I am using zero-inflated Poisson regression.
> Unlike a standard Poisson regression, neither  -estat gof- nor
> -poisgof- commands report the goodness of fit for a zero-inflated
> Poisson.
> How may I find the goodness of fit for zero-inflated Poisson regression?

*
*   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/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index