PAPP101 - s07: Life Tables (I)

Summary

This session has introduced life tables in their classical demographic form, with examples mostly of mortality, but describing their wider use in any area involving duration data.

It has given some brief examples of the use of life tables in ascertaining risk of surviving from any age or duration time to any older age or longer duration time. As such the use of life tables is invaluable in a wide range of studies.

We have described how the single most useful column is that known as the l_x column and that that column alone can be used to derive the complete table. Also that the l_x column, a synthetic cohort, is common to survival analysis generally and may be all that is necessary for many analyses.

To construct the life table thus far we have seen how central death (or failure) rates can be converted to risks (probabilities of failure) and then the _nl_x column simply constructed, via the _nq_x and _np_x columns.

So we have explored the life table up to the lx column, which for many purposes is quite sufficient. How to take the life table further than the lx column to arrive at other useful columns, including the calculation of life expectancy, is the subject of the next session.

What we have not done is explore how life tables can also be constructed from survey data. This is the subject of a session in Module 3.