What is bias?

What is bias?

One of the most important things within the AP Stats course is the foundation of valid and reputable data; without it, there is no basis on which to formulate a solid conclusion. 

In AP Statistics, bias means a systematic tendency for a method to produce values that are consistently too high, too low, or otherwise unrepresentative of the truth. In Unit 3, this usually refers to problems in sampling, surveys, or measurements that make data less representative of the population. Later in the course, AP Statistics also uses the term biased or unbiased to describe point estimates. In AP Statistics, bias matters because a biased sample or survey method can produce results that are not representative of the population, making any conclusions less trustworthy.

There are several important sources of bias in AP Statistics. Four common ones are explained below, but you should also watch for poorly worded or leading questions, which can create response bias.

more resources to help you study

cheatsheetsscore calculatorkey terms

1. Response Bias 🙋‍♀️

• Response bias occurs when the wording of a question or the situation in which it is asked influences respondents to answer inaccurately or untruthfully. This means the responses may not reflect the respondents’ true opinions or behaviors.
• For instance, if an environmental science teacher asks his or her class students who recycles, the students are much less likely to answer truthfully to this figure of authority.
• This type of bias can severely skew data towards one response over others.

2. Voluntary Response Sample ✋

• A voluntary response sample occurs when people choose for themselves whether to participate in a survey or poll. This sampling method often produces biased results because people with strong opinions are more likely to respond.
• Another common example is a call-in poll on television. Since participation is completely voluntary, the responses are likely to come from individuals who feel strongly about the topic.
• As a result, the sample is likely to be biased and may not represent the population accurately.

3. Nonresponse Bias ❌

• Nonresponse bias occurs when individuals selected for the sample do not respond, and those who do respond differ in an important way from those who do not respond.
• The people who do respond do not represent the entire population, so nonresponse bias can be detrimental to the results.

4. Undercoverage Bias 📞

• Undercoverage occurs when some groups in the population are left out of the process of choosing the sample or have a smaller chance of being selected than others.
• For example, if a surveyor wanted to poll voters for an upcoming election and used a phone book to contact residents in his area, his survey would suffer from undercoverage bias.
• There are many reasons that undercoverage is present--one reason is that young adults (think 18-26) do not typically have landlines of their own.
• Another reason is that the extremely wealthy and extremely poor populations are not as likely to be listed in the phone book; all three of these groups were undercovered, so the results will show definite bias.

Another source of bias is measurement bias. This happens when a measuring instrument or process systematically records values that are too high or too low. For example, a miscalibrated scale that always adds 2 pounds produces biased measurements.

Additional Resources

• Introduction into Stats: Do the Data We Collected Tell the Truth? 
• Watch a video review about 🎥 Sampling Methods and Sources of Bias