A nonresponse bias question starts with people selected for a survey but fewer people actually answering. Compare the group named in the conclusion with the group whose responses produced the percentage. People who answer may feel differently from people who don't, even if the initial selection was random. A large number of responses can't remove that risk.
Hints
- Hint 1
The target population is the whole group the team wants to describe. Compare all current subscribers with the group whose answers were counted. Are those groups necessarily alike?
- Hint 2
A random sample means chance chooses who receives the questionnaire, not who replies. Could willingness to pay affect who completes it?
Step-by-step
Compare invitees with respondents
Step 1Identify the group in the conclusion
No Desmos needed. This asks who the survey can represent, not for a calculation. The target population is the whole group the team wants to describe. Here, that's all current subscribers, so the result needs to reflect that group.
- Step 2
Identify whose answers produced the percentage
The team randomly selected 20,000 subscribers, but only 1,900 responded. Those 1,900 are the respondents, the people whose answers were counted. The reported 70% comes from the 1,330 respondents who said yes, not from all 20,000 people selected.
- Step 3
Explain why the replies may not represent everyone
This creates a risk of nonresponse bias: subscribers willing to pay the extra $5 might be more likely to answer than those unwilling to pay. Randomly selecting invitees doesn't guarantee representative replies. So 70% may describe the respondents without accurately describing all current subscribers. Choice C.