On Wed, 26 Feb 2003 15:14:18 +0100 Stefan Englund
<[email protected]> wrote:
> Dear Statalisters
>
> I want to do a Survival analysis with multiple episodes and with time
> varying covariates. The data set looks something like the following
>
> id year event x1 x2
> 1 1975 0 0,86 0,97
> 1 1976 0 0,11 0,97
> 1 1977 1 0,51 0,79
> 1 1978 0 0,62 0,88
> etc . . . .
> 1 1990 0 0,75 0,85
> 1 1991 1 0,75 0,80
> 1 1992 0 0,73 0,82
> etc . . . .
> 1 1995 0 0,80 0,85
> 2 1975 0 0,90 0,56
> 2 1976 1 0,80 0,64
> 2 1977 0 0,80 0,50
> etc . . . .
> 2 1995 0 0,90 0,67
> 3 1975 0 0,45 0,91
> etc . . . .
> 3 1980 1 0,93 0,70
> 3 1981 0 0,91 0,62
> etc . . . .
> 3 1995 0 0,99 0,50
> etc
> etc
> 18 1995 0 0,33 0,21
>
> The X:s are varying every year and the events are cuts in the unemployment
> benefit. The problem I have is to set the appropriate duration in stset. I
> want the spell to be from the year after the last cut up to the present
> cut. Am I totally wrong or can I do this using Stata? Can someone please
> help me?
