Last Update: March 15, 2011

Errors in measurements and comparisons

Count comparisons are subjects to many sources of errors which are known in epidemiology as different forms of bias (bias meaning that one does not measure a true value). Such biases can, in general, be categorized as selection bias (due to the way the study subjects are selected), information bias (due to the way the information in the study is obtained), and screening bias (due to the peculiarities of screening for early manifestations of disease, which brings along the possibility of lead-time bias, length-bias, and overdiagnosis).

Background publications

Morabia A (2004) Epidemiology: An epistemological perspective. In History of epidemiological methods and concepts , Morabia A (ed) pp 1-126. Birkhäuser: Basel , section 3.7

Vineis P (2002) History of bias. In History of epidemiological methods and concepts , Morabia A (ed) Birkhäuser: Basel , pp. 327-336. Older version in SPM.

Key papers

• Berkson J (1946). Limitations of the application of fourfold table analysis to hospital data. Biometrics 2:47-53. PDF
  • Greenland S. Commentary from Evolution of Epidemiologic Ideas (ERI 1987) Berkson J "Limitations of the application of fourfold table analysis to hospital data." PDF
• Copeland KT, Checkoway H, McMichael AJ, Holbrook RH (1977) Bias due to misclassification in the estimation of relative risk. Am J Epidemiol 105 : 488-495 PDF
• Feinstein AR, Sosin DM, Wells CK (1985). The Wil Rodgers phenomenon. Stage migration and new diagnostic techniques as a source of misleading statistics for survival in cancer. N Engl J Med 312: 1604–1608.
• Fox AJ, Collier PF (1976). Low mortality rates in industrial cohort studies due to selection for work and survival in the industry. Br J Prev Soc Med 30: 225–230. PDF
• Glesby MJ, Hoover DR (1996). Survivor treatment selection bias in observational studies: examples from the AIDS literature. Ann Intern Med 124: 999–1005. PDF
• Greenland S (1996). Basic methods for sensitivity analysis of biases. Int J Epidemiol 25:1107-1116. PDF
• Guy WA (1843). Contributions to a knowledge of the influence of employment upon health. J Roy Stat Soc 6: 197–211. PDF
• Hernán MA, Hernández-Díaz S, and Robins JM (2004). A structural approach to selection bias. Epidemiology 15(5):615-625. PDF
• Hutchinson GB, Shapiro S (1968). Lead time gained by diagnostic screening for breast cancer. J Natl Cancer Inst 41: 665–681.
• Kraus AS (1954). The use of hospital data in studying the association between a characteristic and a disease. Public Health Rep 69: 1211–1214. PDF
• Medical Research Council, Report of the Patulin Clinical Trials Committee (2004). Clinical Trial of Patulin in the Common Cold. Int J Epidemiol 33:243-246. PDF
  • Commentary: Kaptchuk TJ, and Kerr CE (2004). Commentary: Unbiased divination, unbiased evidence, and the patulin clinical trial. Int J Epidemiol 33:247-251. PDF
  • Commentary: Sutherland I (2004). Commentary: MRC Patulin trial. Int J Epidemiol 33:252. PDF
  • Commentary: Chalmers I, and Clarke M (2004). Commentary: The 1944 patulin trial: the first properly controlled multicentre trial conducted under the aegis of the British Medical Research Council. Int J Epidemiol 32:253-260. PDF
• Newell DJ (1962). Errors in the interpretation of errors in epidemiology. Am J Public Health 52: 1925–1928. PDF
• Roberts RS, Spitzer WO, Delmore T, Sackett DL (1978). An empirical demonstration of Berkson's bias. J Chronic Dis 31: 119–128. PDF
• Sackett DL (1979). Bias in analytic research. J Chron Dis 32: 51–63. PDF
• Silber ALM, Horwitz RL (1986). Detection bias and relation of benign breast disease to breast cancer. Lancet i: 638–640. PDF