AP U.S. Government topic quiz4 questions

4.6 Evaluating Public Opinion Data

Choose the best answer, check your reasoning, and use the explanation to correct any gap.

Question 1

Question 1 of 4

An election-night exit poll samples voters leaving precincts on Election Day but does not cover early or mail voters, who cast 45 percent of the ballots. A report applies the estimate to all voters. Which limitation is most direct?

Early and mail voters belong to the population named in the report but never enter the exit-poll frame, creating undercoverage. Refusals among approached precinct voters would be nonresponse, and interviewing a sample creates sampling variation, but the stated design gap is the omitted voting modes.
Show answer and explanation

C. Undercoverage because early and mail voters could not enter the exit-poll sample

Early and mail voters belong to the population named in the report but never enter the exit-poll frame, creating undercoverage. Refusals among approached precinct voters would be nonresponse, and interviewing a sample creates sampling variation, but the stated design gap is the omitted voting modes.

Question 2

Question 2 of 4

Two independent surveys using the same sound design estimate support for a river-restoration bond at 45 percent and 49 percent. Each estimate has a reported 95 percent margin of error of plus or minus four percentage points. Which conclusion is most defensible?

Both independent estimates contribute uncertainty to their difference. With comparable 95 percent margins, an approximate margin for the difference is √(4²+4²)≈ 5.7 points, larger than the observed four-point gap. Sampling variation can therefore explain the apparent movement; a difference-specific calculation, not comparison with one poll's margin, supports that conclusion.
Show answer and explanation

D. Movement remains uncertain because both independent estimates contribute sampling uncertainty to their difference

Both independent estimates contribute uncertainty to their difference. With comparable 95 percent margins, an approximate margin for the difference is √(4²+4²)≈ 5.7 points, larger than the observed four-point gap. Sampling variation can therefore explain the apparent movement; a difference-specific calculation, not comparison with one poll's margin, supports that conclusion.

Question 3

Question 3 of 4

Comparable sampled residents are randomly assigned to answer the same recycling-compliance question either with a live interviewer or in an anonymous web form. Completion is high in both groups. Reported compliance is higher with the interviewer, while administrative records are closer to the anonymous estimate. Which explanation best fits?

Random assignment and high completion make respondent composition less persuasive as an explanation. The socially approved answer is more common with an interviewer, while the anonymous estimate better matches the records, supporting social-desirability response bias rather than coverage, order, or panel effects.
Show answer and explanation

A. Interviewer presence encouraged socially desirable reports of compliance

Random assignment and high completion make respondent composition less persuasive as an explanation. The socially approved answer is more common with an interviewer, while the anonymous estimate better matches the records, supporting social-desirability response bias rather than coverage, order, or panel effects.

Question 4

Question 4 of 4

A workplace survey samples employees correctly, but night-shift workers respond at a lower rate. The analyst weights completed cases until night-shift and day-shift shares match the workforce. Scheduling preferences differ sharply by shift. What limitation remains?

The weights restore the observed share of completed interviews from each shift, but they do not observe missing employees' preferences. If night-shift respondents differ from night-shift nonrespondents, the weighted estimate can remain biased even though the shift proportions now match.
Show answer and explanation

B. Night-shift respondents may not represent night-shift nonrespondents on scheduling preferences

The weights restore the observed share of completed interviews from each shift, but they do not observe missing employees' preferences. If night-shift respondents differ from night-shift nonrespondents, the weighted estimate can remain biased even though the shift proportions now match.

Quiz complete