selection bias
A problem that results when individuals or groups are improperly chosen to take part in a study. For instance, an overbalance of women from a certain age group. This can spoil the reliability of the study findings, and means that the conclusions may not be sound or apply to the women concerned. Ideally, the subjects in a study should be similar to one another and to the larger population (for example, all individuals with the same disease or condition) from which they are drawn. If there are important differences or similarities in those subjects that have not been disclosed, the results of the study may not be valid.
For more information, see Breastcancer.org's Clinical Trials section.
Audio contributed by: Marisa Weiss, M.D.