In research, it is a source of systematic error that can influence the outcome of the experiment. Bias can be introduced into research by factors such as, among others, nonrandom sampling, faulty subject instructions, or expectations of the researcher or the participants. In a PDD study looking at the validity of blind scoring, for example, a researcher using only cases which were verified by the original examiner is likely biasing the study, since cases in which the original examiner made the wrong diagnosis could be systematically excluded from the research sample. Researchers attempt to control bias through experimental design.
