Which bias is introduced when the comparison group provides a poor estimate of exposure frequency in a case-control study or disease frequency in a cohort study?

Which bias is introduced when the comparison group provides a poor estimate of exposure frequency in a case-control study or disease frequency in a cohort study?

• A. Bias
• B. Recall bias
• C. Selection bias
• D. Information bias

Answer: (C)

Selection bias occurs when the comparison group isn’t representative of the population that produced the cases, or when how subjects are chosen affects the exposure or disease frequencies being measured. In a case-control study, if controls have an exposure frequency that differs from the source population that gave rise to the cases, the estimated association between exposure and disease becomes distorted. In a cohort study, if the comparison group has a different underlying risk or if follow-up is uneven, the observed disease frequency in the two groups is biased. This bias is about who is included and how they’re chosen, not about how well exposure or disease is measured after data collection. By contrast, recall bias involves differences in remembering past exposure, and information bias involves misclassification of exposure or disease data, not the fundamental issue of selecting a non-representative comparison group.

Selection bias occurs when the comparison group isn’t representative of the population that produced the cases, or when how subjects are chosen affects the exposure or disease frequencies being measured. In a case-control study, if controls have an exposure frequency that differs from the source population that gave rise to the cases, the estimated association between exposure and disease becomes distorted. In a cohort study, if the comparison group has a different underlying risk or if follow-up is uneven, the observed disease frequency in the two groups is biased. This bias is about who is included and how they’re chosen, not about how well exposure or disease is measured after data collection. By contrast, recall bias involves differences in remembering past exposure, and information bias involves misclassification of exposure or disease data, not the fundamental issue of selecting a non-representative comparison group.