Scientific calculations - The principles of sampling

Epidemiology is the study of the distribution and patterns of health and disease, in and across populations.

Scientists study the incidence of disease, using medical records, and use medical research to investigate causes of disease.

When investigating health and disease where risk factors are involved, scientists can't study every person on the planet. Scientists must study samples.

Samples must be representative – they must reflect the population under study. They cannot:

  • focus on certain groups and ignore others
  • be affected by bias

To obtain a representative sample:

  • the study must be wide, and cover all groups in the population being studied
  • the study must be random within these groups

Sampling might be systematic in that it is carried out at regular time intervals.

When working with samples of human populations, studies must take account of possible variations owing to differences:

  • between the sexes
  • resulting from people of different ages taking part
  • in people's lifestyles

Analyses can be carried out on sub-sets of data within the whole study, so that comparisons can be made within the same age group, for instance.

In scientific studies, it may be difficult to control lifestyle factors that have the potential to affect the study, particularly if the study is long-term. People involved in a long-term study of the effects of alcohol intake, for instance, will have different lifestyle factors, such as diet and exercise, that could potentially affect the results of the study.

Where possible, scientists adjust data mathematically so that factors can be removed that might distort a comparison between groups. For example, data could be adjusted to make allowances for the effects of age on rates of disease.