Copyright © University of Cambridge. All rights reserved.

## 'Very Old Man' printed from http://nrich.maths.org/

### Why do this problem?

This

problem
provides an excellent scenario for making and testing statistical
hypothesis. It can be attempted at a variety of levels of
statistical sophistication ranging from making sense of the data to
a full statistical analysis.

### Possible approach

There is quite a lot of information in this problem to digest.
Students might need to spend some time reading through the problem
and making sense of the data. It would work best as a homework or
task where students are given time to think about the problem and
then to come up with their own analysis. Students could then
compare their answers at a later time. From these, the concensus
for a 'best' analysis might emerge.

### Key questions

Before starting this problem, how might we organise or
represent the data? Are you clear as to how we might test the
hypothesis?

### Possible extension

This question naturally raises its own extension: students
might try to improve their answers by searching for more data on in
the internet. pose their own questions, make their own hypothesis
or pursue similar ideas through the materials on the Understanding
Uncertainty site.

### Possible support

Suggest students simply make plots of the life expectancy data
and extrapolate these graphs.