Conducting age-period-cohort model?

[Pallop Siewchaisakul]

Dear all

I am interested in analysing the incidence trends of cancer using APC model.

I did read the Age–period–cohort modeling by MJ Rutherford paper but I can't arranged an appropriate data for it. Does anyone have an example dataset for the poprisktime command and how is the master dataset should be for apcfit command?

It seems like the apcfit command will conduct the apc model in yearly interval. Does anyone know how to do the apc model in 5-years interval of age, period and cohort by using Stata.

PS: I do sorry if my English is not good grammar. I do welcome all of your suggestion and comment.

Best regards,
Pallop

[Mark Rutherford]

Dear Pallop,

Apologies for the slow reply to this post - a colleague drew my attention to it.

It is possible to use apcfit or indeed apcspline to fit to 5-year data, rather than the 1-year intervals.

Using poprisktime is not a pre-requisite for running apcfit and you may well have your data in a format that would be suitable. If you already have a variable indicating the population size then this can be used immediately as the offset in the command.

There are example datasets that are released as ancilliary files with the command that should be useful in deciding if your data are in the appropriate form.

Best wishes,

Mark.

[Pallop Siewchaisakul]

Dear Mark

Thank you very much for your suggestion and taking your time to reply to my post.
I will try, so much things to learn on this topic.

Best regards,
Pallop