EBM Consult

Observational Bias in Statistical Analysis


    • Observational bias is a common type of systematic error that occurs in studies where there is inaccurate:
      • Classification of disease
      • Measurement of disease
      • Exposure
    Common Reasons for Observational Bias
    • Detection Bias:
      • Increased monitoring of a group of exposed patients due to a known risk factor compared to an unexposed group that leads to a higher probability of detecting the disease.
    • Observer Bias:
      • Misclassification of date because of differences in observer interpretation or expectations regarding the study.
    • Recall Bias:
      • Relies on the subject to remember certain information accurately.Subjects who have a negative outcome may be more likely to report certain exposures compared to subjects who had a positive outcome. 
    • Reporter Bias:
      • The subject either over or under report exposures due to perceived social stigmatization or based on what that subject believes to be important.
    • Recall bias is most commonly seen in retrospective studies.