Since epidemiologic research deals with populations, epidemiologists are always concerned with how representative the sample they are studying is of the general population. This is because they have to make estimates about disease distribution.

The ability to choose the best estimate goes beyond statistical bias and reliability. Epidemiology addresses the accuracy of estimates from a measurement perspective rather than sampling theory. When estimating the frequency & distribution of health & disease, the focus is on how a variable should be defined and the evaluation of measurement bias.

Accuracy= Bias + Reliability