A thousand answers, an incomplete sample
A large survey does not by itself fix who has been left out.
What we know
- Surveys have uncertainty from sampling, and also errors that do not simply disappear by collecting more answers: coverage, nonresponse or biased questions.
- Fictional scenario: a museum has 60% local visitors and 40% tourists. In random samples of 100 from each group, 40 locals and 90 tourists support a proposal. Assume these percentages represent their groups for this calculation.
How to answer
Choose the conclusion best supported by the facts and press Check. You can revise an incorrect answer and use help whenever you wish.
Preparing the case…