Hello, dear users of the Stata program. Now I'm conducting a reserach on duration of unemployment. I want to use Cox regression to determine the probability of exiting unemployment.
I prepared the file with unemployment spells in accordance with Stata journal article https://www.stata-journal.com/sjpdf....iclenum=dm0029.
I obtained 180 spells the defined the beginning and ending months for them, duration and censoring flag.
Do I need to reduce observations within 1 spell just to get number of observations as number of spells?
Because now I have data for each observation from each spell and I can't merge other data set with individual characteristics (gender, age, etc). I think this would be sensible.
if yes, how should I do this? I could not find such instructions and I can't continue the work.
0 Response to Spell-type data for Cox regression
Post a Comment