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: St: predicted values in zinb regression


From   Nick Cox <[email protected]>
To   "'[email protected]'" <[email protected]>
Subject   RE: st: RE: St: predicted values in zinb regression
Date   Fri, 18 Mar 2011 18:24:50 +0000

Although the details of the models are important, the point is made clearer by contemplation of e.g. 

sysuse auto
gen gpm = 1 / mpg 
twoway scatter gpm weight || lfit gpm weight 

or the results of 

logit foreign weight 

Nick 
[email protected] 


-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of rachel grant
Sent: 18 March 2011 18:13
To: [email protected]
Subject: Re: st: RE: St: predicted values in zinb regression

> Thank you Nick
> That explains why the model also failed to predict the large peaks that I
> also had in the data, although in many cases the model approximated the
> peaks and troughs, they were definately smoothed out.
Rachel Grant
>
> On 18 March 2011 17:51, Nick Cox <[email protected]> wrote:
>>
>> Not surprising at all. Every model bar a perfect model is likely to fail
>> to predict the extreme values observed of the response variable, as the
>> predicted response is a kind of smoothing or averaging. The question is more
>> do you want to regard rounded predictions as, practically, predictions of
>> zero (or other results).
>>
>> Nick
>> [email protected]
>>
>> rachel grant
>>
>> I have used the command predict n to generate a set of predicted Y
>> values after running zinb regression. I have my predicted values but
>> there are no zeros in the predicted values at all, although in most
>> cases the predicted value was very small where a zero should be e.g.
>> 0.35. Is this OK or have I done something wrong? As I had two highly
>> significant predictors of zeros in the model I was expecting zeros in
>> the predicted values.
>>
>> *
>> *   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/
>
>
>
> --
> regards, Rachel
>
>



-- 
regards, Rachel

*
*   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/


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