Bias in Sampling and Surveys

What Is Bias?

Bias occurs when the method of collecting data systematically favors certain outcomes. A biased study does not accurately represent the population.

Types of Sampling Bias

Voluntary Response Bias

People choose whether to participate. Those with strong opinions are more likely to respond.

Example: An online poll asking "Do you support this policy?" — people who feel strongly will disproportionately respond.

Convenience Sampling Bias

The researcher selects individuals who are easy to reach.

Example: Surveying students in the cafeteria to represent all students.

Undercoverage Bias

Some groups in the population are left out of the sampling frame.

Example: A phone survey that only calls landlines misses people who only use cell phones.

Nonresponse Bias

Selected individuals cannot be contacted or refuse to participate.

Example: A mailed survey with only a 20% return rate — the 80% who didn't respond may differ systematically from those who did.

Types of Response Bias

Question Wording Bias

The way a question is phrased can influence the answer.

Example: "Do you agree that the government should protect endangered species?" (leading question)

Social Desirability Bias

Respondents give answers they think are socially acceptable rather than truthful.

Example: "How often do you exercise?" — people tend to overreport.

Interviewer Effect

The presence or characteristics of the interviewer influences responses.

Summary Table

| Type of Bias | Problem | Solution | |-------------|---------|----------| | Voluntary response | Strong opinions overrepresented | Use random sampling | | Convenience | Not representative | Use random sampling | | Undercoverage | Some groups excluded | Improve sampling frame | | Nonresponse | Non-responders differ | Follow up, incentives | | Question wording | Leading questions | Use neutral wording | | Social desirability | Dishonest answers | Anonymous surveys |

AP Tip: When identifying bias, (1) name the type of bias, (2) explain the direction of bias (overestimates or underestimates), and (3) explain why in context.