I am conducting a Weibull Survival analysis (recurrent event) with the time-frame between 1970 and 2010. One question is the difference in survival time between gender. However in the earlier years for 1970-1973 there are no women observed. Since calendar year is not my time scale this shouldn't be a problem right? Everyone in my sample starts at zero (onset of risk).
I did inlcude possible period effects as dummy variables (for example new hollywood era and blockbuster era).