Hi everyone,

I'm trying to estimate a proportional hazard model with a piecewise constant baseline hazard on unemployment-spell level data by using the stsplit command. Failure in this context means the transition from unemployment to employment. As I have multiple spells per individual in the dataset, I want to specify that my data is of multiple-failure type. Additionally, I need to specify that the "clock" resets every spell, i.e. I need the _t0 variable generated by stset to be 0 for each observation (i.e. each spell). (Meaning that, even for the second or third entry for each individual, the analysis time when the record begins is 0)
According to the spell-type "stset"-Guidebook (https://www.stata.com/support/faqs/s...ell-type-data/), this is only possible if I do not specify the ID() option when stsetting the data. However, stsplit does require an ID-variable.
Can someone help me with this issue?
Thanks a lot!

Valentin