Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


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

RE: st: regression where independant var with gaps


From   Nick Cox <n.j.cox@durham.ac.uk>
To   "'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu>
Subject   RE: st: regression where independant var with gaps
Date   Fri, 15 Jun 2012 18:25:23 +0100

If you know that a value is zero, then it is not missing. Situations with genuine missing values are not comparable. 

Zero values are not a barrier to regression unless you want to transform a predictor with logarithms. That doesn't seem to be an issue here. If you have zeros on your response -- and as here your response is counted -- then an appropriate regression method is Poisson regression, to which you have been pointed several times. More generally any generalised linear model with logarithmic link can cope easily with some zero values on the response as long as positive conditional mean is a reasonable specification. 

References to previous pointers in related threads started by you: 

Austin Nichols' post of 13 June 

http://www.stata.com/statalist/archive/2012-06/msg00668.html

My post of 9 June 

http://www.stata.com/statalist/archive/2012-06/msg00508.html

My post of 8 June 

http://www.stata.com/statalist/archive/2012-06/msg00491.html

Nick 
n.j.cox@durham.ac.uk 


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of tashi lama
Sent: 15 June 2012 17:57
To: statalist@hsphsun2.harvard.edu
Subject: RE: st: regression where independant var with gaps

True. Since my data are readership hits, I populated the days without readership hits zero. 

I was thinking somewhere in the line where in a real research, you will have missing values. Say, you are trying to identify the income solely based on education. If you pick 25 ppl for the research and you know the incomes of all 25 ppl but education of only 24 ppl. You would have a missing value unless you decide to ignore it all and take 24 ppl for your research. How would you do regression in this case where you decide to include all the ppl and hence have a missing value. 

 

Thanx 

----------------------------------------
> From: njcoxstata@gmail.com
> To: statalist@hsphsun2.harvard.edu
> Subject: Re: st: regression where independant var with gaps
> Date: Fri, 15 Jun 2012 17:25:51 +0100
>
> You have shown data examples like
> this with values every day but some zeros. Be advised: if the gaps are
> known to contain zeros, then they are not gaps. This is not a matter
> of convention or preference.
>
> Nick
>
>
> On 15 Jun 2012, at 16:54, tashi lama <ltashi32@hotmail.com> wrote:
>
> >
> > Hello all,
> >
> > I have a univariate time series with gaps like following
> >
> >
> >
> > date hits
> >
> > 01/01/2011 40
> >
> > 01/02/2011 20
> >
> > 01/03/2011 9
> >
> > 01/06/2011 6
> >
> > 01/10/2011 2
> >
> > ........
> >
> >
> >
> >
> >
> > Can I run a regression since the my independant vars have gaps?
> *
> * 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/

*
*   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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index