Chapter 6 Quiz: What Is Public Opinion?

Multiple Choice (10 questions)

1. Philip Converse's concept of "non-attitudes" refers to:

a) Opinions held by people who choose not to vote b) Survey responses generated on the spot with no stable underlying evaluative content c) Political views that fall in the ideological center d) Attitudes that cannot be measured by standard survey instruments

Answer: b. Converse found that many respondents fabricated answers to policy questions they had never thought about, generating responses that were essentially random across survey waves.

2. In Zaller's Receive-Accept-Sample (RAS) model, which of the following best describes the "Sample" component?

a) The random selection of survey respondents from the population b) The process by which citizens receive information from political elites c) The mechanism by which citizens draw on accessible considerations to construct a survey response d) The rejection of political arguments inconsistent with predispositions

Answer: c. Zaller's "Sample" refers to respondents sampling from their currently accessible considerations when constructing an expressed opinion — not to statistical sampling.

3. The thermostatic model of public opinion predicts that when government increases spending on a social program, public opinion on that program will:

a) Become more favorable over time as citizens experience the benefits b) Become more supportive of further spending increases c) Shift toward preferring less spending on that program d) Remain stable, because policy doesn't affect opinion

Answer: c. The thermostatic model predicts a corrective feedback: as policy moves in one direction, public demand for further movement in that direction falls. When spending increases, the thermostat cools demand for more spending.

4. Walter Lippmann's 1922 concept of the "pseudo-environment" argues that:

a) Citizens are fundamentally irrational and incapable of political judgment b) Citizens navigate a simplified, stereotyped representation of the world rather than political reality directly c) Public opinion is entirely manufactured by elites and has no grassroots component d) Only direct democracy can overcome the distortions of representative institutions

Answer: b. Lippmann argued that citizens operate through mental pictures of reality shaped by media and social communication — not through direct engagement with political facts.

5. Converse found that, in mass publics, ideological constraint was:

a) Universally present and organized along a coherent left-right dimension b) Present among elites but absent in mass publics, explaining polarization c) Rare — most citizens' issue positions were weakly correlated with one another d) Strongest among low-education respondents who rely on social cues

Answer: c. Converse's central finding was that the ideological coherence apparent among political elites and commentators did not characterize ordinary voters. Issue positions across different policy domains were largely uncorrelated in mass publics.

6. According to the RAS model, a high-information citizen is most likely to:

a) Be more open-minded and persuadable than a low-information citizen b) Accept political arguments on their merits regardless of partisan alignment c) Receive more political communication AND be better at resisting arguments inconsistent with predispositions d) Have fewer accessible considerations, making their opinions more stable

Answer: c. Zaller's model produces the counter-intuitive result that high-information citizens are better at both receiving elite communication and filtering it through partisan predispositions, making them in some ways more resistant to cross-cutting persuasion.

7. The "spiral of silence" refers to:

a) The tendency of polling organizations to suppress unflattering poll results b) The process by which minority opinion becomes less visible as its holders withhold public expression c) The mathematical convergence of poll estimates toward the true population value d) The declining response rates in modern telephone polling

Answer: b. Noelle-Neumann's spiral of silence describes the self-reinforcing suppression of minority opinion: as people perceive their view is in the minority, they express it less, making it appear still more marginal, causing further suppression.

8. When researchers conduct a study and find that 25% of respondents express support for a fictional piece of legislation with a realistic-sounding name, this best illustrates:

a) High levels of political engagement in the sample b) Survey acquiescence bias in agree/disagree questions c) Converse's non-attitudes — respondents generating opinions about issues they have no real views on d) The thermostatic model's prediction of responsive opinion formation

Answer: c. Support for fictitious legislation is the canonical demonstration of non-attitudes: respondents will express opinions about policy objects that don't exist, demonstrating that the survey response has no stable underlying content.

9. Which aggregation problem is most directly illustrated by the fact that polls of "likely voters" systematically produce different results than polls of "all adults"?

a) Acquiescence bias inflates agreement with all propositions b) The choice of who counts as "the public" is a political decision with distributional consequences c) Probability sampling is impossible without a complete population list d) Question order effects distort late responses

Answer: b. The likely voter / all-adults gap illustrates that "public opinion" means different things depending on which public you count. Likely voters are older, whiter, and more educated — so choosing this definition embeds demographic and political assumptions about whose voice matters.

10. Which of the following is the most accurate statement about the relationship between survey questions and public opinion?

a) A well-designed survey question discovers the true underlying opinion of respondents b) Survey questions are contaminated measures of an otherwise objective public opinion c) Survey questions participate in constructing the opinion they purport to measure d) The ideal survey question is one that produces the same result regardless of wording

Answer: c. The chapter's central argument, consistent with Converse, Zaller, and measurement research, is that questions don't discover pre-existing opinions — they help constitute the opinions they measure through their framing, wording, and context.

Short Answer (5 questions)

11. Explain in 100-150 words why Converse's finding about "non-attitudes" is troubling for democratic theory — not just for polling methodology.

Model answer: Democratic theory typically assumes that citizens hold genuine preferences about policy, and that government responsiveness to those preferences constitutes democratic accountability. If Converse is correct that a substantial fraction of survey responses are non-attitudes — responses fabricated on the spot with no stable underlying content — then much of what we call "public opinion" is not a genuine preference signal at all. Government "responsiveness" to these non-attitudes is not democratic accountability; it is responsiveness to noise. Moreover, if politicians can change expressed "preferences" simply by framing issues differently — by making different considerations accessible — then the preferences governments claim to respond to may be substantially manufactured by elites rather than genuinely derived from citizens. This challenges the basic democratic claim that government gets its legitimacy from the expressed will of the governed.

12. Describe the Receive-Accept-Sample model in no more than 150 words. Why does it predict that expressed opinions will vary based on question context?

Model answer: Zaller's RAS model proposes that opinion formation involves three stages. First, citizens receive political communication from elites — more communication if they are more politically engaged. Second, they accept or resist that communication based on their predispositions — accepting arguments consistent with their values and resisting inconsistent ones. Third, when asked a survey question, they sample from the considerations currently accessible in memory and generate an expressed opinion based on that sample. Because accessible considerations vary depending on context — what questions were asked before, what news events have occurred recently, what the question wording makes salient — the opinions respondents express vary accordingly. The opinion is constructed in the moment of answering, not retrieved from a stable storage location. Context determines what's accessible; what's accessible shapes the expressed opinion.

13. What is "ideological constraint," and what did Converse find about its distribution in American mass publics?

Model answer: Ideological constraint refers to the degree to which an individual's positions on different political issues are logically or ideologically connected — the extent to which knowing someone's position on one issue predicts their positions on others. High constraint means positions form a coherent pattern (e.g., consistently liberal or consistently conservative across economic, social, and foreign policy issues). Converse found that ideological constraint was largely absent in mass publics in the 1950s-60s. Most ordinary citizens' positions on different policy issues were weakly correlated; knowing someone's position on government health insurance told you very little about their position on civil liberties or foreign aid. Only a small fraction of the electorate — perhaps 2-4%, whom Converse called "ideologues" — organized their political views around coherent ideological principles.

14. Define social desirability bias. Give one specific example of how it might distort a political poll, and describe one technique researchers use to reduce it.

Model answer: Social desirability bias is the tendency of survey respondents to give answers that conform to perceived social norms rather than their actual attitudes — to answer in ways that make them look good to an interviewer or to themselves. In political polling, a classic example is the overstatement of voter turnout: respondents who did not vote in a recent election will often report that they did, because not voting is perceived as socially irresponsible. This inflates self-reported turnout relative to actual turnout. One technique for reducing social desirability bias is self-administered surveying (web or paper questionnaires) rather than interviewer-administered telephone surveys. The absence of a human interviewer removes the immediate social pressure to seem like a good citizen, and respondents are somewhat more likely to report truthfully on sensitive behaviors and attitudes.

15. What is the "aggregation problem" in public opinion measurement? Use a specific example to illustrate why it matters.

Model answer: The aggregation problem refers to the difficulty of combining individual opinions into a meaningful collective "public opinion" without making contested assumptions about whose voice counts how much. Every aggregation method embeds implicit political choices. For example, a national poll on immigration policy that weights equally to the adult population will produce a different result than one that weights to the likely voter population — because likely voters are older, whiter, and more educated than all adults, and these groups differ in their immigration views. There is no neutral answer to which is "correct": the adult population estimate reflects a more inclusive democratic picture but may poorly predict actual policy influence; the likely voter estimate better predicts electoral consequences but systematically underweights the preferences of groups who face structural barriers to voting. Choosing one over the other is a political decision dressed as a methodological one.