In surveys, how do sampling bias and response bias differ?

• A. Sampling bias occurs when respondents lie; response bias occurs when sample size is too large.
• B. Sampling bias is about data cleaning; response bias is about sampling frame.
• C. Sampling bias occurs when the sample isn't representative; response bias occurs when respondents misreport due to social desirability or recall.
• D. Sampling bias occurs in lab tests; response bias occurs in emails.

Answer: (C)

Sampling bias and response bias are two ways survey results can be distorted. Sampling bias happens when the group chosen for the survey isn’t representative of the whole population you want to learn about, usually because of how participants are selected—some groups are over- or under-represented, pulling results away from what’s true for the population. Response bias occurs when the answers themselves are skewed because respondents misreport or misremember information, often due to social desirability, memory errors, or misunderstanding questions.

For example, surveying only urban adults to infer national healthcare usage illustrates sampling bias because the sample doesn’t reflect the broader population. Conversely, people underreporting smoking on a health questionnaire to appear more healthful shows response bias, since the data are distorted by how individuals respond rather than by who was surveyed.