1. 10 Public Opinion and Political Socialization
Pollsters use Google search results to track interest in the 2012 presidential candidates, Governor Mitt Romney and President Barack Obama.
Zhang Jun/Newscom

2. 10
Learning Objectives
Trace the development of modern public opinion research.
10.1
Describe the methods for conducting and analyzing different types of public opinion polls.
10.2
The Founders of the country recognized that "all government rests on public opinion." In this chapter, we'll learn the process of political socialization, which describes how Americans form their values and beliefs about politics. Pollsters and politicians take frequent polls to measure what Americans think about policies, even though most Americans have little knowledge about complex political matters.
Still, as we we'll learn in this chapter, public opinion has a critical influence on what politicians do and say.

3. 10 Learning Objectives Assess the potential shortcomings of polling.
10.3
Analyze the process by which people form political opinions.
10.4

4. 10
Learning Objectives
Evaluate the effects of public opinion on politics.
10.5

5. Roots of Public Opinion Research
10.1
Roots of Public Opinion Research
The Earliest Public Opinion Research
The Gallup Organization
The American National Election Studies
Public opinion is what the public thinks about an issue or a particular set of issues. Polls are used to estimate public opinion.
Almost since the beginning of the United States, various attempts have been made to influence public opinion about particular issues or to sway elections. Presidential polling has been around since 1916, and George Gallup was the first to use scientific polling methods to determine public opinion.

6. The Earliest Public Opinion Research
10.1
The Earliest Public Opinion Research
Polling to predict winner of elections
Polling to discover public opinions
Walter Lippmann's Public Opinion
Literary Digest
Straw poll
Sample
As early as 1824, one Pennsylvania newspaper tried to predict the winner of that year's presidential contest, showing Andrew Jackson leading over John Quincy Adams. But public opinion research did not really emerge as a science until the 1930s, following the publication of Walter Lippmann's book Public Opinion in As political scientists gained interest in public opinion in politics and other walks of life, different methods to capture it were tried. The survey was the primary method.
The popular magazine Literary Digest was a pioneer in the use of the straw poll, an unscientific survey used to gauge public opinion, with initial success. That success ran out, however, when the magazine in 1936 predicted the defeat of President Franklin D. Roosevelt, who actually won in a landslide. It later became clear that the upper middle class had been overrepresented in the sample of Americans the magazine had surveyed.

7. 10.1 Is polling always accurate? Bettmann/Corbis
Not only did advance polls in 1948 predict the Republican nominee Thomas E. Dewey would defeat Democratic incumbent President Harry S Truman, but on the basis of early and incomplete vote tallies, some newspapers' early editions published the day after the election declared Dewey the winner. In this photo, a triumphant Truman holds aloft the Chicago Daily Tribune.
Bettmann/Corbis

8. The Gallup Organization
10.1
The Gallup Organization
George Gallup
Dissertation on how to measure the readership of newspapers
Expanded research to study public opinion about politics
Increased use of polling to market products and candidates
Newspapers—like the Chicago Daily Tribune held by President Harry S Truman with the famous "Dewey Defeats Truman" headline that got the 1948 presidential election predictions wrong—probably wished they'd hired a pollster named George Gallup. Gallup wrote his dissertation on how to measure the readership of newspapers at the University of Iowa. He later expanded his research to include public opinion about politics.
Soon businesses and politicians were using polling to discover what the market wanted. The Gallup Organization continues to predict the winners of presidential elections.

9. 10.1 FIGURE 10.1 How successful has the Gallup Poll been?
As seen here, Gallup's final predictions have been remarkably accurate. Furthermore, in each of the years in which a significant discrepancy exists between Gallup's prediction and the election's outcome, a prominent third candidate factored in. In 1948, Strom Thurmond ran on the Dixiecrat ticket; in 1980, John Anderson ran as the American Independent Party candidate; in 1992, Ross Perot ran as an independent.
Sources: Marty Baumann, "How One Polling Firm Stacks Up," USA Today (October 27, 1992): 13A; 1996 data from Mike Mokrzycki, "Pre-election Polls' Accuracy Varied," Atlanta Journal and Constitution (November 8, 1996): A12; 2000 data from Gallup Organization, "Poll Releases," November 7, 2000; 2004, 2008, and 2012 data from USA Today and CNN/Gallup Tracking Poll,

10. The American National Election Studies
10.1
The American National Election Studies
Focuses on attitudes of the electorate
How voters voted
Party affiliation
Opinions of parties and candidates
Surveys before and after mid-term and presidential elections
Efforts to measure public opinion have benefited from social science surveys, such as the American National Election Studies, or ANES, which is funded largely by the U.S. government through the National Science Foundation.
ANES asks citizens how they voted, what their party affiliation is, and their opinions of parties and candidates. The ANES conducts surveys before and after mid-term and presidential elections, and has compiled long-term studies of the electorate.

11. 10.1 Which term describes the people who are polled?
Survey
Straw poll
Sample
Registered voters
Let's take a minute to answer this review question about polls.

12. 10.1 Which term describes the people who are polled?
Survey
Straw poll
Sample
Registered voters
As pollsters learned in the 1936 election, it is important to make the sample for a survey as representative of the whole population as possible.

13. Conducting and Analyzing Public Opinion Polls
10.2
Conducting and Analyzing Public Opinion Polls
Designing the Survey and Sample
Contacting Respondents
Analyzing the Data
Those who conduct polls must first determine what questions they want answered and how to phrase those questions. Then they must determine the sample, or subset, of the group whose attitudes they wish to measure. Finally, they have to figure out how to contact respondents. The different types of polls include telephone polls, in-person interviews, and Internet polls. And once the results are in, they must be analyzed.

14. Designing the Survey and Sample
10.2
Designing the Survey and Sample
Determining the content and phrasing of the questions
Wording is crucial
"Push polls"
Selecting the sample
Population
Random sampling
Stratified sampling
Before pollsters can ask anyone anything, they must ask themselves a number of critical questions first. When it comes to the content and phrasing of the poll, special care must be taken in constructing the question; differently phrased questions could lead to different answers.
Especially in politics, sometimes the intent of the poll is to skew the results by negatively phrasing the question. These are called push polls, and they might ask a respondent: "If you knew Candidate X beat his wife, would you vote for him?" Reputable pollsters avoid push polls.
Once the questions are selected, pollsters must determine the population whose attitudes they want to gauge. If it's simply the population at large, pollsters might use a random sampling. If they're looking at specifics in income, family size, or ethnicity, they might use stratified sampling, which is based on U.S. Census data that provide the number of residences in an area and their location.

15. Contacting Respondents
10.2
Contacting Respondents
Telephone polls
Random digit dialing survey
Tracking Polls
Despite the increase in cell phones and decrease in landlines, telephone polling is still the chief form of polling used in surveys, especially during campaign season. The most common type of telephone poll is the random digit dialing survey, in which a computer selects a phone number for dialing.
Sometimes campaigns use tracking polls, in which a small sample of people are polled every 24 hours. These allow campaigns to measure short-term developments and the effects of certain campaign strategies.

16. How are polls conducted?
10.2
How are polls conducted?
The most common method of conducting a public opinion poll is via telephone. These polls are often administered by survey researchers working at large phone banks, such as the one shown here.
Melanie Stetson Freeman/The Christian Science Monitor via Getty Images

17. 10.2 FIGURE 10.2 What does a daily tracking poll look like?
Day-to-day fluctuations in public opinion on electoral contests are often shown through tracking polls. This figure shows the ups and downs of the 2012 presidential election. President Barack Obama led for much of the race, although the polling data generally remained within the margin of error. Note, particularly, Mitt Romney's gains after the first presidential debate.
Source: USA Today and CNN/Gallup Poll results, aspx?ref=interactive.

18. Contacting Respondents
10.2
Contacting Respondents
In-person interviews
Exit polls
Internet polls
Scientific polls versus unscientific Web polls
Some polls, however, such as the American National Election Studies, use in-person interviews. The most common of these types of polls are the exit polls that are conducted as citizens emerge from voting on Election Day.
Finally, there are Internet polls. Some polling companies have devised scientific strategies to ensure that the results of Internet polling are legitimate, but they are not to be confused with so-called Web polls that allow anyone to weigh in on a topic.

19. 10.2
Analyzing the Data
Reveals implications for public policy and political campaigns
Data analyzed by computers
Subgroups of population, such as men versus women, age groups, or political ideology, may be analyzed
Reporting the data
News organizations or campaigns
Data gathered from the poll must be analyzed for meaning. This step reveals the implications of the data for public policy or for political campaigns. The data are crunched by computers, which may further break down the information by subgroups. This allows campaigns to see how their candidate is faring among women versus men, or in a certain age group.
Once the data have been analyzed, the results are generally reported. This could be done by the media or directly by a campaign.

20. 10.2 Which type of poll is designed to give a skewed result?
Push poll
Straw poll
Tracking poll
Stratified sampling
Here's another review question about polling.

21. 10.2 Which type of poll is designed to give a skewed result?
Push poll
Straw poll
Tracking poll
Stratified sampling
Push polls are intended to give some negative information about a candidate's opponent to push people away from voting for the opponent and toward the other candidate.

22. Shortcomings of Polling
10.3
Shortcomings of Polling
Survey Error
Limited Response Options
Lack of Information
Difficulty Measuring Intensity
Lack of Interest in Political Issues
Polls may have several shortcomings that create inaccuracies. These include:
•survey errors
•not having enough respondent options to reflect public opinion on an issue
•polling those who lack the information necessary to accurately respond
•the inability to measure the intensity of public opinion on an issue
•the public's lack of interest in political issues

23. Survey Error 10.3 Margin of error Sampling error
Natural errors in statistical measurement
Sampling error
Happens when a pollster draws an improper sample
Because polls are based on a sample that's intended to reflect the general public, the results are close but not exact. Polls allow for a margin of error. Typically, the margin of error in a sample of 1,000 people is plus or minus about 4%. So if 52% of the people in a sample of 1,000 say they plan to vote for Candidate X, pollsters can predict that between 48 to 56 people support this candidate.
Another type of survey error is the sampling error. This occurs when pollsters fail to sample all elements of a population. For example, the poor and the homeless are generally underrepresented in polls.

24. Limited Respondent Options
10.3
Limited Respondent Options
Yes/No (Approve/Disapprove)
May not give respondents sufficient room to answer
"Feeling thermometer"
Respondents rate feelings 0—100
There are problems with limited response options in polls. Respondents who feel very strongly about an issue or are somewhat neutral about it may not be able to adequately—or accurately—respond to a poll that offers only yes/no or approve/disapprove options.
For that reason, some pollsters—including the American National Election Studies—use a feeling thermometer style of question, in which respondents rate from 0 to 100 their feeling on a given prompt.

25. Lack of Information 10.3 Filter questions
Gauge how much respondents know about or have thought about an issue
Responses to personal questions/moral value questions
Feelings stronger about some issues
Depending on the issue, respondents may not have enough background information about an issue to accurately answer poll questions. In these cases, especially if the poll is about a complex policy or budgetary issues, the pollster may first ask a filter question to determine how much a respondent knows or has thought about the issue. Based on this response, up to 20% of respondents could be excluded from the rest of the poll.
Most people have opinions about personal or moral issues, such as drugs, crime, or abortion. When asked questions about these areas, respondents give fewer "no opinion" or "don't know" responses.

26. Difficulty Measuring Intensity
10.3
Difficulty Measuring Intensity
Issues with strong feelings, such as:
Big government
Death penalty
Support for the war on terrorism
Smaller issues, such as:
Electoral College
Absentee ballot laws
Another shortcoming of polls is that they can't measure the intensity of a respondent's feelings. A person might say yes to a question about absentee ballot laws but not care that much. On the other hand, a person who responds to a question about the war on terror may have very strong feelings.

27. 10.3 Can polls measure intensity of opinion?
One of the greatest shortcomings of most public opinion polls is that they measure direction of public opinion, but not intensity. Here, members of Westboro Baptist Church demonstrate intense opposition to homosexuality by protesting outside the Supreme Court.
Chip Somodevilla/Getty Images

28. Lack of Interest in Political Issues
10.3
Lack of Interest in Political Issues
Many people lack an opinion on certain policies.
Policies that do not affect people directly
Policies that do not involve moral values
Foreign policy
Less likely to generate interest
Domestic policy
More likely to generate interest
Many people have a difficult time forming an opinion on policies that don't appear to affect them directly or aren't connected to a moral value question. This is especially true with Americans and foreign policy. Domestic policies, such as those affecting health care, bank bailouts and employment are more likely to generate Americans' interest.

29. 10.3
10.3 Which of these can be a problem in designing, executing, or analyzing the results of a poll?
Lack of interest in political issues
Difficulty in measuring intensity
Limited response options
Survey error
All of the above
Polls aren't perfect. Can you answer this review question about polling problems?

30. 10.3
10.3 Which of these can be a problem in designing, executing, or analyzing the results of a poll?
Lack of interest in political issues
Difficulty in measuring intensity
Limited response options
Survey error
All of the above
All of the responses can be a problem when designing, executing, and analyzing the results of a poll.