Introduction

Most science is comparative. Researchers often need to know if a particular experimental treatment has had an effect, or if there are differences among a particular variable measured at several different locations. For example, does a new drug affect blood pressure, does a diet high in vitamin C reduce the risk of liver cancer in humans, or is there a relationship between vegetation cover and the population density of rabbits? But when you make these sorts of comparisons, any differences among treatments or among areas sampled may be real or they may simply be the sort of variation that occurs by chance among samples from the same population.

Here is an example using blood pressure. A biomedical scientist was interested in seeing if the newly synthesised drug ‘Arterolin B’ had any effect on blood pressure in humans. A group of six humans had their systolic blood pressure measured before and after administration of a dose of Arterolin B. The average systolic blood pressure was 118.33 mm Hg before and 128.83 mm Hg after being given the drug (Table 5.1).

The average change in blood pressure from before to after administration of the drug is quite large (an increase of 10.5 mm Hg), but by looking at the data you can see there is a lot of variation among individuals – blood pressure went up in three cases, down in two, and stayed the same for the remaining person.