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: nbreg or ologit?
From
"Shittu, Aminu" <[email protected]>
To
"[email protected]" <[email protected]>
Subject
Re: st: nbreg or ologit?
Date
Wed, 20 Jun 2012 01:48:56 -0700 (PDT)
Thanks - Joerg!
Aminu.
________________________________
From: Joerg Luedicke <[email protected]>
To: [email protected]
Sent: Tuesday, June 19, 2012 10:32:27 PM
Subject: Re: st: nbreg or ologit?
On Tue, Jun 19, 2012 at 12:23 PM, Shittu, Aminu <[email protected]> wrote:
> In a count data from individual animals (20 samples collected from each of the 23 farms and production questionnaire administered); should the count data be modelled using "nbreg" or the count response be converted to an ordinal variable as low, medium and high egg counts and modelled using "ologit?"
No. (Whether -nbreg- is the right thing cannot be said based on the
information you provide. But reducing a count variable to a
categorical one makes no sense whatsoever.)
> Could both be interpreted similarly?
With a wide definition of "similar", regression coefficients from all
kinds of models can be interpreted similarly.
> Furthermore, is it ok including farm as random effect as animals are clustered within farms?
I assume you mean varying intercepts across farms. Sure, why not? You
could perhaps include varying coefficients as well, if it seems
reasonable.
J.
>
> Regards,
>
> Aminu.
>
> *
> * 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/