Duration Variable for Survival Analysis in the UCDP Peace Agreement Dateset

[Hajo Benning]

Deal All!

its my first time doing a survival analysis and I want to apply it on the durability on Peace Agreements. Does anyone know how to create the duration variable (Time to Event variable), that is, counting the years from signing till failure of the agreement?

Thanks!

hajo

[Carlo Lazzaro]

Hajo:
welcome to the list.
I would take a look at -st- entries in Stata .pdf manual, to start with.
The mandatory textbook for those dealing with survival analysis in Stata is: http://www.stata.com/bookstore/survi...-introduction/

[Hajo Benning]

Dear Carlo,
I always have the feeling that the explanations in the stata pdf etc beginn one step after my problem: before declaring the data to be survival data I have to create this time-to-event variable, which is difficult for me. I have two time variables: "pa_date" as date of signing of the agreement and "duration" as date for the end of the agreement. "DyVi05" is my failure variable coded as 1 if violence occurs within 5 years between the same parties; 0 otherwise. CID is my conflict ID. So far I got this, but not sure if right....:

sort CID Year
bysort CID: replace DyVi05=. if DyVi05==1
replace DyVi05=. if CID==10 & Year>=1995
sort CID
by CID: replace DyVi05=1 if DyVi05[_n-1]==0 & DyVi05==.
by CID: gen dur_fail = _n (this is my time-to-event variable)
replace dur_fail =. if DyVi05==.
drop if DyVi05==.

Thanks for any help!
Hajo

[Carlo Lazzaro]

Hajo:
please act on the FAQ about posting your code within CODE delimiters. Thanks.
You can find inspiration in the following Stata dataset:

use http://www.stata-press.com/data/r14/drugtr.dta

[Marcos Almeida]

Carlo gave excellent advice.

As a side note, since you said you are just starting to delve into survival analysis, maybe a nice start-up could be watching Stata's You-Tube-Channel series on the matter, presented by Chuck Huber.