well, i have a question about the analysis of duration models..Ok my data
are not time-varying, but i have one observation for each id (country)..
The problem is not the duration variable, but the covariates i have
In particular, i have a variable of various random dates, one date for each
country which indicates the year which this country was colonized(which these dates are different from the t0 or t1 of the duration variable)...Lets say i want to estimate a cox-ph model..how i treat this