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: Model for Poisson-shaped distribution but with non-count data
From
Nick Cox <[email protected]>
To
[email protected]
Subject
Re: st: Model for Poisson-shaped distribution but with non-count data
Date
Tue, 6 Dec 2011 09:00:48 +0000
-gammafit- (SSC) has been available for some years, but random
intercepts are fancier than it does.
However, I am more concerned with two dogmas surfacing here without
little or no foundation, that
1. Poisson models are for counts only
2. You choose models based on the marginal distribution of the
response or outcome variable.
See http://blog.stata.com/2011/08/22/use-poisson-rather-than-regress-tell-a-friend/
for an excellent exposition that makes no such assumption on Poisson.
On the evidence here I would still try out -poisson- or a related
command.
I don't know where #2 comes from. Every decent modelling text
explains that assumptions are about conditional distributions, and not
that important even then.
Nick
On Tue, Dec 6, 2011 at 5:09 AM, Paul Millar <[email protected]> wrote:
> An appropriate distribution would be gamma, but that is only available
> for panel data. Anyone up for creating a new MLE command?
On Mon, Dec 5, 2011 at 6:34 PM, Owen Gallupe <[email protected]> wrote:
>> Does anyone know what type of regression model I should use? I've been
>> searching and have not been able to find a modeling approach designed
>> to meet the distributional properties of a variable I am hoping to
>> analyse.
>>
>> The dependent variable has what looks like a Poisson distribution, but
>> with non-count data. About 28% of the sample scores somewhere between
>> 0 and 1. The highest value is 182.6. Skew = 2.256; kurtosis = 10.002.
>> N=2776.
>>
>> I have tried bootstrapped linear regressions and linear regressions
>> after employing a normalizing transformation using lnskew0 (though the
>> normalization is not perfect and results in a bimodal residual
>> distribution).
>>
>> One further complication is that I need to include random intercepts.
*
* 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/