0
$\begingroup$

suppose I have an output of survival probabilities over time as follows:

Jan: .987
Feb: .962
Mar: .943
Apr: .894

What is the correct way to calculate conditional survival probability? For example, the probability of surviving to February given you have survived through January? Is it simply .962/.987 = .975?
This number is close to my observed value which is good, but it starts to get less accurate over the rest of the months.

Is March going to be .943/.975=.967, and april is going to be ..894/.967? Those numbers dont seem to be accurate to me and the further I get out, the more inaccruate the probabilities become

$\endgroup$

1 Answer 1

1
$\begingroup$

Actually, the probability for March given Feb is 0.943/0.962 or about 0.980 So Survival(March) = Survival(Jan) * S(F|J) * S(M|F) = 0.987 * 0.975 * 0.980 which comes out to the 0.943 provided

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.