History of bias

Summary

Epidemiologists have always been conscious of the importance of controlling for distortions, although the definition itself of bias has changed over time. Central to this discussions in the past was the relative vulnerability of different study designs to bias and uncontrollable confounding (confounding being clearly distinguishable from bias, as a problem of inter-mixed causal effects due to the non-random distribution of risk factors within the study population). In particular, controversy arose over aspects of case-control study design. Also a formulation of “typologies of bias” during the 1970s helped to define some of the most important sources of distortion in the design, analysis and interpretation of epidemiological studies. The subsequent period — until now — has been characterised by more formal and systematic definitions.