1
$\begingroup$

I'm running a negative binomial regression. I found big differences in the results* if I compute with the glm y x, (nbinomial 1) link(log) and glm y x, (nbinomial ml) link(log) (for Stata) command.

What is the difference between computing with ML** and with one?

I have big differences in the significance level and I would strongly prefer to take the model with nbinomial 1. The differences between the AIC values are 200 (lower values for nbinomial ml).

*The results showed big differences in the significance and standard error level. ** Stata state's that the "Negative binomial parameter estimated via ML and treated as fixed once estimated". It seems that the parameters are estimated via Maximum Likelihood but what does this mean for the other model?

$\endgroup$
6
  • $\begingroup$ So the differences between SPSS and Stata are for when the negative binomial dispersion term is set to 1? And when estimating the dispersion they are the same? (Note to others, both SPSS and Stata as a default use the NB2 type of dispersion, so I don't believe that is the problem.) $\endgroup$
    – Andy W
    Mar 17, 2015 at 18:12
  • $\begingroup$ Best to post the code for each, I've checked results a few times using nbreg y x, dispersion(mean) with SPSS's GENLIN Y WITH X /MODEL X DISTRIBUTION=NEGBIN(MLE) LINK=LOG., and they have always been exactly the same (although I have had situations in which Stata converges where SPSS does not). As always, are you sure you are estimating the same models? (Same variables, same records with no missing data?) SPSS and Stata have potential differences in how an exposure/offset is specified as well. $\endgroup$
    – Andy W
    Mar 17, 2015 at 18:17
  • $\begingroup$ Thanks fot the answer Andy. Yes the results are the same with SPSS and Stata. The differences are between the glm y x, (nbinomial 1) link(log) and glm y x, (nbinomial ml) link(log). What exactly is the difference between model nbinomial ml and nbinomial 1? And which model is to take? As I mentionned I have big differences in the significance level and I would strongly prefer to take the model with nbinomial 1. The differences between the AIC values are 200 (Better values for nbinomial ml). $\endgroup$
    – Walter
    Mar 18, 2015 at 8:27
  • 1
    $\begingroup$ Try to update the question with this info. As written it is confusing, and suggests differences between the results in SPSS vs Stata, not different models within Stata. (nbinomial 1) means you set the dispersion to an arbitrary constant, 1. I highly doubt setting it to a constant is appropriate, but one hypothetical situation I can imagine is if by past results you have strong reason to believe it is equal to 1. $\endgroup$
    – Andy W
    Mar 18, 2015 at 11:31
  • $\begingroup$ Dear Andy. Thanks for the response. I updated my question. I haven't this hypothetical situation. Normally one does it with ML? $\endgroup$
    – Walter
    Mar 18, 2015 at 13:18

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.