Skip to content
AP Statistics · Exploring One-Variable Data and Collecting Data

Survey design

Survey design appears in 39 reviewed AP Statistics questions and sits in Exploring One-Variable Data and Collecting Data. Below are worked examples showing exactly how the exam tests it, with the answer marked and every distractor explained.

How Survey design is tested

Example 1Medium
A company surveys employees to understand job satisfaction. The survey uses a 10-question instrument, and at the end it asks employees to submit their responses in a sealed envelope to be opened by an external analyst. What feature of this design reduces what type of bias?
  1. AConfidentiality reduces response bias, since employees answer more honestly when they know supervisors cannot identify their individual answers.Correct
  2. BConfidentiality prevents undercoverage by ensuring every department in the company is included in the sample.
  3. CConfidentiality reduces voluntary response bias by making participation in the survey mandatory for all employees.
  4. DConfidentiality eliminates nonresponse bias by encouraging every selected employee to return a completed survey, so no one is left out of the results.

Why A is right

When employees know their responses are confidential (not traceable to them individually), they are more likely to answer honestly about dissatisfaction or criticism of management without fear of retaliation. This reduces social desirability bias — the tendency to give favorable or politically safe answers — which is a type of response bias.
Why the other 3 choices miss
B
Incorrect because undercoverage is about missing groups in the sampling frame, not about whether responses are confidential. Confidentiality addresses accuracy of responses, not who is included in the sample.
C
Incorrect because confidentiality does not make participation mandatory. Voluntary response bias involves self-selection into an open survey; confidentiality affects answer honesty, not whether participation is voluntary.
D
Incorrect because confidentiality may increase response rates but does not eliminate nonresponse bias (some people still won't respond). Nonresponse bias depends on who doesn't respond, and confidentiality doesn't guarantee all non-responders will participate.
Example 2Harder
A survey asks respondents: 'How many times have you driven after consuming alcohol in the past year?' The average response is 0.3 times. Researchers suspect the true average is much higher. Which of the following survey design modifications would most likely reduce the bias in responses?
  1. AUsing a randomized response technique that provides plausible deniability for sensitive answersCorrect
  2. BConducting the survey by telephone instead of in person
  3. CChanging the question to 'Do you drink and drive? (Yes/No)'
  4. DIncreasing the sample size from 500 to 5,000

Why A is right

A randomized response technique introduces a random mechanism (like a coin flip) that determines whether a respondent answers the sensitive question or a harmless one. Because the researcher cannot determine which question any individual answered, respondents have plausible deniability, reducing social desirability bias and encouraging more truthful responses to sensitive questions.
Why the other 3 choices miss
B
Incorrect. Telephone surveys may actually increase social desirability bias compared to anonymous written surveys because respondents are speaking to another person. The mode change alone does not address the fundamental bias.
C
Incorrect. Converting to a yes/no question loses information (frequency) and may actually increase social desirability bias because the dichotomous format forces an admission of a stigmatized behavior.
D
Incorrect. Increasing sample size reduces random error but does not correct systematic bias. If respondents underreport at all sample sizes, a larger sample will still produce a biased estimate.
Example 3Easier
A newspaper wants to estimate the proportion of adults in a city who support a proposed tax increase. A reporter stands outside the city library on a Tuesday morning and asks the first 100 adults who walk by. This sampling method is:
  1. AA simple random sample of city adults
  2. BA voluntary response sample
  3. CA convenience sample that is likely biasedCorrect
  4. DA stratified sample based on location

Why C is right

This is a convenience sample because the reporter selects individuals who happen to be accessible at a particular time and place. People at a library on a Tuesday morning are not representative of all adults in the city — they may be more educated, older, retired, or have different employment patterns, introducing bias.
Why the other 3 choices miss
A
Incorrect because a simple random sample requires every adult in the city to have an equal chance of being selected, which is not the case here.
B
Incorrect because the adults did not choose to participate; the reporter approached them. Voluntary response involves people self-selecting into the sample.
D
Incorrect because stratified sampling involves dividing the population into groups and sampling from each; selecting people at one location does not constitute stratification.

Studied alongside

Other concepts from Exploring One-Variable Data and Collecting Data.