Bias is introduced into scientific research in a variety of ways, not least through the individuality of the researcher, her likes and dislikes in modes of working, physical location, research methods, her social background and values, and a host of other possible influential factors. Bias can arise is the design of the investigation, such as sampling errors. Type I and Type II errors are those of omission and commission, respectively, where relevant parameters or influences have been ignored, on the one hand, or not excluded from the experiment on the other. The scientific method has proved invaluable in reducing certain types of bias, but is much weaker in eliminating others.
In the case of medical research, there is concern that studies may be unbalanced because those studied do not constitute a representative sample of the population, and yet the results are used to prescribe treatment for the population as a whole. Medical research on heart disease, for example, tends to be based on younger people, and yet it is amongst older people than its incidence is greatest. An age cut-off is applied which may also reflect the relevance of such studies to women.