
Lecture 8
Estimation of the rate of
mortality
The data we have to work with is not always obtained by direct obser-
vation of cohorts. In the next lecture we will discuss somewhat “im-
perfect” data, that have a more complicated structure. However, in this
lecture we will continue to consider a cohort, i.e. continue to consider
T
1
,... ,T
n
as independent and identically distributed random variables.
As always, denote by F the common distribution function of each T
i
.
Suppose F has density f , and let µ be the rate of mortality (see (1.13)),
µ(x) =
f (x)
1 −F(x)
.
If F is unknown, then µ is also unknown. Let us examine how we can
estimate the rate of mortality ...