Susanna,
I just ran the code below on simulated data and found the dispersion
parameter estimate on the last row in the Parameter Estimates table.
The name of the parameter is "(Negative binomial)".
GENLIN y WITH x1
/MODEL x1 INTERCEPT=YES
DISTRIBUTION=NEGBIN(MLE) LINK=LOG
/PRINT CPS DESCRIPTIVES MODELINFO FIT SUMMARY SOLUTION .
Ryan
On Thu, Mar 3, 2011 at 4:54 PM, SK <
[hidden email]> wrote:
> Hi,
>
> I need to calculate the value of overdispersion parameter from my data. I
> have used the method of moment by using predicted values and residuals from
> Poisson regression. For some reason I don't get same result as the reference
> calculations I have.
>
> I noticed that Genlin -procedure estimates ML-estimates for the
> overdispersion parameter but I didn't find any way to get it printed.
>
> Any ideas how I can get the parameter value?
>
> Thanks,
>
> Susanna
>
> --
> View this message in context:
http://spssx-discussion.1045642.n5.nabble.com/How-to-calculate-the-value-of-negative-binomial-distribution-overdispersion-parameter-tp3408688p3408688.html> Sent from the SPSSX Discussion mailing list archive at Nabble.com.
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
>
[hidden email] (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD
>
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
