Discrete Hazard Model; Where to Begin?

[Marcus Galeste]

Hi All:

I have been out of the loop in consistent use of STATA for a few years and feel I am having to re-learn some seemingly basic steps. I am using a multi-wave dataset (4 waves; wave 1 is baseline) and intend on running a discrete hazard model revolving around a failure event. My dependent variable is dichotomous. I have this variable for each wave with varying occurrences. My questions:

1) what do I need to do with my data to ensure it is in the correct format to eventually run a discrete hazard?
2) what "time" variable is required? I was planning on using each wave (3, 9, 15 months), but am struggling with how to transform the data so that it is in the correct format to run analyses.

Any assistance or direction to a place to start would be greatly appreciated!

[Roman Mostazir]

I think here is where to start as there are several key concepts:

http://www.stata.com/manuals13/ststset.pdf

[Steve Samuels]

See the chapter "Discrete" in the ST manual. It's page 19 for Version 14.1. For more detail, see Lessons 3 & 6 on Stephen Jenkins's fine web page "Survival analysis with Stata", link below, and his book draft on the same page. What do you mean by a "variable for each wave with varying occurrences" ?

[Marcus Galeste]

Thank you both so much for your reply! This is definitely a great start.

I suppose "variable for each wave with varying occurrences" was not the best way to describe what I mean. I simply meant that I have this dependent variable for each wave.