STATA Negative Binomial Regression results differ from SPSS

[doriangray]

SPSS is providing me expected results (significance of covariates) in a NBR Model. But I am also running the model in Stata and the results are different. Some covariates are significant across both with common valence (-/+), but some are not. I have already fixed the "parameterization of dispersion" to constant, but the results are still different. Any thoughts on what I am missing?

 

[hlsmith]

Are the differences in effect sizes trivial or not, referencing significance isn't informative since it is an all or nothing classification. Please post outputs!

 

[doriangray]

Are the differences in effect sizes trivial or not, referencing significance isn't informative since it is an all or nothing classification. Please post outputs!

These are the output files from the same variables, same data, run in Stata and SPSS. @hlsmith let me know if any other information is required