Re: st: R-squared measures proposed by Cameron and Windmeijer (1996) in stata

[Nick Cox <njcoxstata@gmail.com>]

You are limiting yourself to replies from those who know the paper in
question or are willing to go and read it in order to be able to
answer your question. Thus far, that appears to be an empty set.

Nick

On Sat, Mar 26, 2011 at 11:23 AM, Francisco Rowe <frowe@ucn.cl> wrote:
> Regarding question 1, I did something, but I would like to know if I am alright.
> I ran a NBRM2 using nbreg. Then with fixed alpha parameter, I ran it again using the glm command in order to get the deviance of that model. After, I applied the same procedure for a model with the variables in my model.  Thus, using the deviances for these two model with only constant and variables, I calculated a Pseudo R-squared as 1-(Deviance for model with variables/Deviance for the only-constant model).
>
> Is this the way how this measure of fit should be calculated in Stata?
>
> I would appreciate your comments.
>
> On 25/03/2011, at 2:46 PM, Francisco Rowe wrote:
>
>> Hi,
>>
>> I have two questions regarding the estimation of a Negative Binomial Regression Model with quadratic variance (NBRM2).
>>
>> 1) How could the R-squared for NBR2 based on deviances proposed in the paper below be implemented using STATA? Does it require any changes if robust standard errors are used for the estimation?
>> Cameron, C and Windmeijer, F 1996, 'R-squared measures for count data regression models with applications to health-care utilization', Journal of Business & Economic Statistics, vol. 14, no. 2, pp. 209-20.
>>
>> 2) In general -for what I have seen-, why measures of fit in form of (Pseudo) R-squared are relatively low (<0.3)? Are there any special conditions that cause this?
>>
>> I would appreciate your comments.

*
*   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/