I am confused about how the emprical variance is being calculated in the solution:
The parts of the question that are relevant to my question are:
Ok so in the solution, it says that since there is no truncation or censoring, Greenwood's approximation is equal to the empirical variance (I understand this part). So it says let d be the number of deaths in each year. Then,
[(2d/1000)((1000-2d)/1000)]/1000 = .00016
What I don't understand is how is the left side of the equation equal to the empirical variance? Can anybody walk me through how we arrive at the left side of the equation?
The parts of the question that are relevant to my question are:
Quote:
From a 2 year mortality study of 1000 lives beginning at exact age 40, you are given: (i) observed deaths are distributed uniformly over the interval (40,42] (ii)Greenwood's approximation of Var(S(2)|{n40,n41})=.00016 |
Ok so in the solution, it says that since there is no truncation or censoring, Greenwood's approximation is equal to the empirical variance (I understand this part). So it says let d be the number of deaths in each year. Then,
[(2d/1000)((1000-2d)/1000)]/1000 = .00016
What I don't understand is how is the left side of the equation equal to the empirical variance? Can anybody walk me through how we arrive at the left side of the equation?
ASM 14th ed, 25.5
0 commentaires:
Enregistrer un commentaire