AP U.S. Government and Politics · Lesson 2

Measuring and judging public opinion

Know Know it

Public opinion is what people think about issues, leaders and policies. The main tool for measuring it is a scientific poll. A good poll draws a random sample, in which every member of the population has an equal chance of being picked, so the sample tends to look like the whole population. Pollsters often weight the results so groups match their real share of the population.

Every sample poll has sampling error. The margin of error gives the range above and below a result in which the true figure probably falls. A random sample of about 1,000 adults has a margin of error near plus or minus 3 percentage points. A bigger sample shrinks the margin, but slowly: it takes about four times as many people to cut it in half.

Pollsters use different kinds of polls for different jobs. A benchmark poll measures a candidate's standing at the start of a campaign. A tracking poll asks the same question of new samples over time to show change. Entrance and exit polls question voters as they arrive at or leave the polling place. A focus group is a small guided discussion that gives depth but does not stand for the whole population.

To judge poll data, ask who was asked, how many, when and how. Watch for question wording that pushes an answer, question order that primes people, samples where people pick themselves (such as call-in or online click polls), and nonresponse bias, when the people who do not answer differ from those who do. A push poll is not a real poll at all: a campaign uses loaded questions to spread negative claims about a rival.

Public opinion matters because officials and candidates watch it. Polls can shape which issues candidates stress, and a candidate rising in the polls may attract extra support, a bandwagon effect. Polls can also miss when it is hard to tell who will actually vote.

Kinds of polls
KindWhat it doesWhen it is used
Benchmark pollMeasures where a candidate stands before the campaign gets goingAt the start of a campaign
Tracking pollRepeats the same question with new samples to show changeAcross a campaign or a term in office
Entrance or exit pollQuestions voters arriving at or leaving the pollsOn primary or election day
Focus groupA small guided discussion that probes feelings and reasonsWhen testing messages or ads

Trap A lead inside the margin is not a real lead

If one candidate leads 47 to 45 and the margin of error is plus or minus 3 points, each figure could be several points higher or lower. The race is too close to call from that poll, so do not pick an answer that calls one side the clear leader.

Worked example Easy

A polling firm wants its survey of 1,000 adults to stand for the whole country. What must it do when choosing whom to ask?

  1. Give every adult an equal chance of being chosenAnswer
  2. Ask only people who volunteer on its website
  3. Interview shoppers leaving one large mall
  4. Choose only people who voted in the last election

How to solve it

  1. The goal is a sample that looks like all adults.
  2. Only random selection gives every adult the same chance.

Why each choice is right or wrong

  • A. Correct. A random sample gives every adult an equal chance to be picked, so the sample tends to match the population.
  • B. Incorrect. Volunteers on a website pick themselves, so they are not a random sample of all adults.
  • C. Incorrect. Shoppers at one mall come from one place and one kind of outing, so they do not stand for the country.
  • D. Incorrect. Recent voters leave out every adult who did not vote, so they cannot stand for all adults.

Explain Explain it

Task verbs: evaluate, explain

To evaluate poll data, say whether you can trust it for the claim being made, then give the reason in terms of the sample, the wording, the timing or the margin of error. A claim about a whole population needs a random sample of that population.

Worked example Medium

A poll asks, "Do you support the governor's reckless plan to waste money on new highways?" What is the main problem with this question?

  1. The sample it uses is too large
  2. Its wording pushes people toward one answerAnswer
  3. It is really an exit poll
  4. It asks about a state issue instead of a national one

How to solve it

  1. Read the question as a respondent would.
  2. Spot the slanted words: reckless and waste.
  3. Loaded words bias the answers, a question wording problem.

Why each choice is right or wrong

  • A. Incorrect. Nothing is said about the sample size, and a large random sample is a strength, not a problem.
  • B. Correct. Words such as reckless and waste push respondents to say no, so the result reflects the wording more than opinion.
  • C. Incorrect. An exit poll questions voters leaving the polls, and this question is not tied to an election day.
  • D. Incorrect. Polls can measure opinion on state issues as well as national ones, so the topic is not the flaw.

Evaluate Your turn

A radio host asks listeners to call in, and 80 percent of the 500 callers say a new city tax is a bad idea. Evaluate whether this result tells you what the whole city thinks, and give the reason.

Write it, then compare. Nothing you type here is saved or graded.

Compare with a model answer

A full answer does this

  • Says the result does not show what the whole city thinks
  • Gives an accurate reason the callers do not stand for the city

Model answer

No, the result does not tell us what the whole city thinks. The callers chose themselves instead of being picked at random, and only people who listen to that show could take part, so the 500 callers may be far angrier about the tax than the city as a whole.

Apply Apply it

Practice these topics. AP Practice questions on exactly what this lesson covers, with an explanation for every answer.