Which approaches are commonly used to mitigate nonresponse bias?

• A. Increasing sample size
• B. Ignoring nonresponse
• C. Using only online surveys
• D. Follow-ups, weighting, imputation, incentives

Answer: (D)

Mitigating nonresponse bias involves addressing who participates as well as what happens to those who don’t. Nonresponse bias occurs when respondents differ in meaningful ways from nonrespondents, so the results don’t represent the whole population. The best approach combines steps that raise the response rate with methods that adjust for remaining differences. Following up with nonrespondents—through several contact attempts, different times, or alternate modes—helps capture groups that might otherwise be underrepresented. Weighting adjusts the survey data to match the population on known characteristics or estimated response propensities, so underrepresented groups get appropriate influence in the estimates. Imputation fills in missing data for cases with partial or missing information, allowing analyses to use all available data while reflecting uncertainty through methods like multiple imputation. Incentives can stimulate participation across diverse groups, further reducing systematic differences between respondents and nonrespondents. Taken together, these strategies directly tackle nonresponse bias rather than leaving it unaddressed or relying on a single data collection method.

Increasing sample size reduces sampling error but does not eliminate bias from who chooses to respond. Ignoring nonresponse leaves the bias intact. Using only online surveys can worsen bias if certain groups are less likely to respond online.