Explain selection bias in polling samples and how to mitigate it.

• A. Bias arises from non-random selection; mitigate with random sampling, weighting, and careful recruitment.
• B. Bias arises from using too many questions; mitigated by shortening the survey.
• C. Bias cannot be mitigated once the poll is conducted.
• D. Bias arises from non-random selection; mitigate with random sampling, weighting, and careful recruitment.

Answer: (D)

Selection bias in polling samples happens when the people who end up in the poll aren’t representative of the whole population because the way the sample is chosen favors certain groups. If certain ages, regions, or socioeconomic groups are more likely to respond, the poll’s results can tilt toward the views of those groups, giving a misleading picture of public opinion.

The best way to counter this is to use probability-based, random sampling so that every individual has a known, nonzero chance of being selected. This reduces the systematic differences that come from how people are included or excluded. In addition, weighting helps place the final results on the same footing as the population by adjusting the influence of respondents to match known population characteristics (like age, gender, race, or education). This corrects for imbalances that creeped in during fielding.

Careful recruitment and outreach are also crucial. Making repeated contact attempts, using multiple modes of data collection (phone, web, in person), and offering incentives can reduce nonresponse bias, where those who don’t respond differ in relevant ways from those who do. Sometimes, combining these approaches with stratified or targeted sampling helps ensure that subgroups are adequately represented.

Keep in mind that simply increasing the number of respondents doesn’t fix selection bias—the issue is about who is included in the sample, not just how many people are surveyed. Mitigation focuses on how the sample is drawn and how the results are adjusted to reflect the broader population.