Statistics / Unit 5: Collecting Data
5.3 Bias in Sampling
Notes
Bias means the method systematically favors certain outcomes, so the statistic tends to miss the parameter in one direction. A larger sample does not fix bias.
- Undercoverage: some groups in the population have no chance of being selected.
- Voluntary response bias: people who choose to respond often have strong opinions.
- Nonresponse bias: many chosen individuals cannot be reached or refuse.
- Response bias: people give inaccurate answers, often because of leading or confusing wording, or the wish to please.
- Example: A call-in poll about a new law mostly hears from people with strong feelings.
Practice
A radio station asks listeners to call in to vote on a question. The main source of bias is:
Only listeners with strong opinions are likely to call.
A phone survey is conducted at 2 p.m. on weekdays using landline numbers only. Which type of bias is most likely?
People without landlines or who are away at work have little chance of being included.
A survey asks, "Don't you agree that the awful cafeteria food should be improved?" This question most likely causes:
The wording pushes respondents toward a particular answer.
Ready for more? Take the Unit 5 test.