Nonresponse bias in public opinion polls is addressed by which practice?

• A. They weight responses to known demographics
• B. They use follow-ups to contact nonrespondents
• C. They increase the sampling error
• D. They publish only positive results

Answer: (A)

Nonresponse bias happens when those who participate in a poll differ from those who don’t, and those differences can distort the overall results. The most effective way to address this in analysis is to weight responses so the sample matches known population demographics like age, gender, education, and region. By applying these weights, each respondent stands in for a larger number of people, compensating for unequal response rates across groups and making the survey’s estimates more representative of the whole population.

Follow-ups to contact nonrespondents can help raise response rates and reduce the potential bias, but they don’t by themselves fix the bias in the data unless the resulting sample aligns with population characteristics. Increasing sampling error would only make estimates less precise, not solve the bias problem, and publishing only positive results ignores the issue entirely.