[Joao Santos Silva]

Dear Steven YH Chen

If you mean predictions, I do not think you can get any meaningful predictions, unless T is large.

Best wishes,

Joao

[Steven YH Chen]

Dear Joao Santos Silva: I am curious whether margins could work too? Could you share some articles with me on why predict or margin could not work after xtpoisson model? My sample size is around 800. Or any post estimation I can run for that?

I see you mentioned before,

If you are using a fixed effects estimator, -margins- will not give you any interesting results. Also, you cannot do predictions. The reason is simple: all of these depend on fixed effects and we do not have consistent estimates for them. Setting the FE at some value is meaningless because the result will depend on the way you measure your regressors. For example, if in the model in #1 we replace age with year of birth the results of -margins- and predictions will change.

In short, do not do it! In the Poisson case the coefficients have a natural interpretation and that should be enough.

However, I don't know why.

Thank you so much!!!


Steven

[Joao Santos Silva]

Dear Steven YH Chen

Maybe this will help.

Best wishes,

Joao

[Steven YH Chen]

Thank you so much Joao Santos Silva !!!

[Patrick Lang]

Joao Santos Silva I understand that the margins command after xtpoisson with fixed-effects does not report reliable results. Is there any other way to create interaction plots without using margins and marginsplot for poisson fixed-effects regression?
Thanks for your help.

Best wishes,
Patrick

[Joao Santos Silva]

Dear Patrick Lang,

I am afraid I have no experience with such plots. All I can do to try to help is to suggest that you check this.

Best wishes,

Joao

[Patrick Lang]

Many thanks for the help Joao Santos Silva .