Distinguish selection bias from information bias in a cohort study and provide an example.

• A. Selection bias occurs when the cohort is selected in a way related to exposure or outcome (e.g., healthier workers are more likely to be included).
• B. Information bias occurs from misclassification of exposure or outcome.
• C. Confounding occurs when a variable is not measured.
• D. Recall bias is the same as selection bias.

Answer: (A)

Selection bias happens when the way you form the cohort influences who ends up in the study in relation to exposure or outcome, creating a non-representative group and biasing the measured association. If healthier workers are more likely to participate, the cohort may underrepresent higher-risk individuals, making the observed link between exposure and disease appear weaker or stronger than it truly is.

Information bias, in contrast, comes from errors in measuring exposure or outcome after participants are enrolled. For example, if smoking status is self-reported and some smokers are misclassified as non-smokers, or if disease diagnosis relies on inconsistent criteria, the exposure–outcome relationship can be distorted.

So, the example given illustrates selection bias—the cohort’s composition is related to health status at enrollment—whereas information bias would arise from mismeasurement during data collection. Confounding is a separate issue involving a third variable related to both exposure and outcome, and recall bias is a form of information bias tied to differential memory of past exposure.