Dear all,
Thank you for the interesting discussion. I am writing a reply because I am encountering a situation that seems to contradict what is said above. It could be that I am making a mistake, but I would like to know your opinion. I am running stcox for a country-year setting, and it turns out some of my subjects fail at time 0 (entry time).
As the documentation on stset says now, "Observations with t = time ≤ 0 are ignored because information before becoming at risk is irrelevant."
According to this, the subjects that fail at time of entry should not be considered by the model, yet they are, according to the standard e(sample) test.
Did stata change its procedure on this? It turns out, moreover, that taking into account failure at entry makes sense from my theoretical perspective, but I am afraid this would be forcing the model.
I would appreciate any ideas of why this might be happening! Thank you!
Thank you for the interesting discussion. I am writing a reply because I am encountering a situation that seems to contradict what is said above. It could be that I am making a mistake, but I would like to know your opinion. I am running stcox for a country-year setting, and it turns out some of my subjects fail at time 0 (entry time).
As the documentation on stset says now, "Observations with t = time ≤ 0 are ignored because information before becoming at risk is irrelevant."
According to this, the subjects that fail at time of entry should not be considered by the model, yet they are, according to the standard e(sample) test.
Did stata change its procedure on this? It turns out, moreover, that taking into account failure at entry makes sense from my theoretical perspective, but I am afraid this would be forcing the model.
I would appreciate any ideas of why this might be happening! Thank you!
Comment