R help - Aug 2007 - extracting dispersion parameter from quasipoisson lmer model

Hi,
  I would like to obtain the dispersion parameter for a quasipoisson model for
later use in calculating QAIC values for model comparison.Can anyone suggest a
method of how to go about doing this?
   
  The idea I have now is that I could use the residual deviance divided by the
residual degrees of freedom to obtain the dispersion parameter. The residual
deviance is available in the summary output for the lmer model. The problem with
this idea is that the residual degrees of freedom is not directly available. I
have been assuming that the estimated/approximate DF can be obtained through the
formula listed below, and then subsequently used to obtain the dispersion
parameter as described above. Is this correct?
   
  samp.m4 <- mcmcsamp(m4, n = 1000)
print(summary(samp.m4))
(eDF <- mean(samp.m4[,"deviance"]) - deviance(m4, REML=FALSE)) 
  # potentially useful approximate DF?

   
  However, rather than going through this roundabout, there appears to be an
easier way to obtain the dispersion parameter. I have noted that the
'quasibinomial' produces a dispersion parameter in the model output, but
the 'quasipoisson' output does not contain this useful number. Given
that each of my models is fit by lmer with a quasipoisson distribution, the
program must be internally calculating the dispersion parameter as it runs.
Perhaps there is a way to obtain this number directly since I assume it has
already been calculated? Can somone who has experience with this code provide
some advice on whether this is possible and how I might manage to do this?
   
   
  Thank you for your time,
   
  Wayne Hallstrom
  403-370-3832
   
   

       
---------------------------------

	[[alternative HTML version deleted]]

Hello,

I have the exact same question: how to extract dispersion parameter from
lmer() quasipoisson model?

I did a search here and saw this post. However, I do not see answer. Could
somebody please help?

Thank you very much,
Julia


wayne hallstrom wrote:
> 
> Hi,
>   I would like to obtain the dispersion parameter for a quasipoisson model
> for later use in calculating QAIC values for model comparison.Can anyone
> suggest a method of how to go about doing this?
>    
>   The idea I have now is that I could use the residual deviance divided by
> the residual degrees of freedom to obtain the dispersion parameter. The
> residual deviance is available in the summary output for the lmer model.
> The problem with this idea is that the residual degrees of freedom is not
> directly available. I have been assuming that the estimated/approximate DF
> can be obtained through the formula listed below, and then subsequently
> used to obtain the dispersion parameter as described above. Is this
> correct?
>    
>   samp.m4 <- mcmcsamp(m4, n = 1000)
> print(summary(samp.m4))
> (eDF <- mean(samp.m4[,"deviance"]) - deviance(m4, REML=FALSE))
>   # potentially useful approximate DF?
> 
>    
>   However, rather than going through this roundabout, there appears to be
> an easier way to obtain the dispersion parameter. I have noted that the
> 'quasibinomial' produces a dispersion parameter in the model
output, but
> the 'quasipoisson' output does not contain this useful number.
Given that
> each of my models is fit by lmer with a quasipoisson distribution, the
> program must be internally calculating the dispersion parameter as it
> runs. Perhaps there is a way to obtain this number directly since I assume
> it has already been calculated? Can somone who has experience with this
> code provide some advice on whether this is possible and how I might
> manage to do this?
>    
>    
>   Thank you for your time,
>    
>   Wayne Hallstrom
>   403-370-3832
>    
>    
> 
>        
> ---------------------------------
> 
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
> 
> 
-- 
View this message in context:
http://www.nabble.com/extracting-dispersion-parameter-from-quasipoisson-lmer-model-tp11992221p22441798.html
Sent from the R help mailing list archive at Nabble.com.