Define selection bias and give an example in public health research.

• A. Selection bias occurs when study participants are not representative due to non-random selection; example: surveying only clinic patients may miss those who do not seek care.
• B. Bias due to measurement errors in outcome.
• C. Bias due to randomization.
• D. Bias due to confounding.

Answer: (B)

Selection bias is a systematic distortion that happens when the way participants are chosen or retained in a study makes the group studied unrepresentative of the target population, leading to biased estimates of associations or effects. For example, surveying only people who visit clinics to study a disease will miss those who do not seek care, so the results may not reflect what’s happening in the broader population.

This idea is what best captures selection bias: non-random selection creates a sample that differs in important ways from those not selected. In contrast, measurement bias arises from errors in how outcomes or exposures are measured, bias from randomization isn’t a standard bias (randomization aims to prevent bias), and confounding involves a third variable related to both exposure and outcome rather than issues with how participants were selected.