Results insignificant when using nbreg

[Zorodzai Murinye]

I have industry-country-year observations for 150 countries for 33 years. My explanatory variable is a dummy variable to indicate when event X occurred. My pre-event period is 10 years and so is the pre-event period. My data follows a poisson distribution but due to overdispersion I used the nbreg command in my regression models instead of poisson.
I ran the command:
nbreg Y X, vce(robust) irr
My results show significance only for the period 3 year pre-event and 2 years post-event. Overall, the F-test for the model is significant.
What could these results be telling me?

[Matt Warkentin]

Hi Zorodzai,

I think you'll find you will receive better responses if you provide more information. You can use the -dataex- command to share a snippet of your data, or you could simply provide more of your code and/or the results, so we can use that information to provide a better response. You can place code in delimiters by selecting the # icon when submitting a response on this website. You can also take screenshots of your results and upload them as an image. Generally, I think I am clear on your data, but I am not clear on your analysis. It sounds like you have panel data which could be analyzed more appropriately using something like -xtnbreg-. If you provide more details I would be happy to try and help provide insights.