POLS 1140
Political Knowledge and Information
Updated Mar 9, 2025
Monday
Overview
Monday:
Recap discussion of Ideology and Issues
General overview of political knowledge
Begin discussion of Jerit, Barabas and Bolsen (2006)
Wednesday:
Alternative conceptions of Weaver, Prowse and Piston (2019)
A1 Section 1: 5-6 pm: Zoom: https://brown.zoom.us/j/97622927036
Thursday:
- A1 Section 2: 5-6 pm: Zoom: https://brown.zoom.us/j/98683021396
Friday:
- Misinformation (Jerit and Zhao 2020)
Revised Groups
Assignment 1:
- Before section:
- Get group contact info
- Brainstorm and share ideas
- Section:
- Chance to brainstorm
- 1 Page Memo:
- General topic
- Elevator pitch
- Outcomes of interest
- Key predictors
Polling the Election
Some recent NYT/Sienna State Polls
- Harris +4 in PA (Sept. 11-16)
- Trump +5 in AZ; +4 in GA; +2 in NC (Sept 17-21)
For Wednesday, please also read: Accurate Polls Hinge on a Tricky Question: Who’s Actually Going to Vote?
Review
What have we learned so far?
Converse (1964)
Ansolabehere, Rodden, and Snyder (2008)
Freeder, Lenz and Turney (2019)
Converse (1964)
Most citizens lack what stable, coherent ideological belief systems
Shallow understanding of liberal-conservative labels
Inconsistent beliefs across issue, particular among the mass public
Centrality of groups and possiblity of issue publics
Ansolabehere, Rodden, and Snyder (2008)
A critique of Converse (1964) suggesting the apparent instabiltiy and incoherence of issue attitudes is a product of measurement error
Using multiple items to measure latent concepts:
- Reduces measurement error
- Increases stability and predictive validity of measures
Freeder, Lenz, and Turney (2019)
A critique of Ansolabehere’s critique, suggesting that simply constructing multi-item scales reduces multiple sources of error and instability, only some of which is classic measurement error
Returns to Converse’s concept of “what goes with what,” using candidate and party placements (how) as a proxy for this form of knowledge.
Show that knowledge of what goes with what is the driving force of ideological consistency and stability
- But only for folks that agree with their party/candidate’s position.
Wednesday
Plan for Wednesday:
Announcements, assignments, review
Finish Jerit et al.
Begin Weave Prowse and Piston
For Friday:
- Read Jerit and and Zhao
Revised Groups
Assignment 1:
- Before section:
- Get group contact info
- Brainstorm and share ideas
- Section:
- Chance to brainstorm
- 1 Page Memo:
- General topic
- Elevator pitch
- Outcomes of interest
- Key predictors
Sections
- First section today from 5-6 pm
- Second section tomorrow from 5-6 pm
First Reflection Papers Due October 12
Due October 12th
May be on any topic we’ve covered so far
- Ideology
- Political Knowledge
- Political Cognition
- Democratic Choice
Additional Readings
Some additional suggested papers
- Mondak (1983)“Public opinion and heuristic processing of source cues”
- Kam (2005) “Who Toes the Party Line? Cues, Values, and Individual Differences”
- Taber and Lodge (2006) “Motivated Skepticism in the Evaluation of Political Beliefs”
- Bartels (2005)“Homer Gets a Tax Cut: Inequality and Public Policy in the American Mind”
- Fowler and Hall (2018) “Do Shark Attacks Influence Presidential Elections? Reassessing a Prominent Finding on Voter Competence”
- Busby et al (2016) “The Political Relevance of Irrelevant Events”
- Miller et al. (2016) “Conspiracy Endorsement as Motivated Reasoning: The Moderating Roles of Political Knowledge and Trust”
Concepts you should be applying
- Measures of Association
- Measures of Uncertainty
- Models of Causation
Questions you should be asking
- What’s the research question
- What’s the theoretical framework
- What’s the empirical design
- What are the results
- What are the conclusions
What’s the research question
What is this paper about?
If you had to summarize this paper in one to two sentences, how would you do it?
What’s the theoretical framework
What’s the argument?
What are the key concepts?
Why do we care?
What’s the contribution?
What are the expectations?
What’s the empirical design
How will the paper convince you of its claims
Is it an experimental or observational design
What data does the paper use
What methods the paper apply
What are the results
What evidence does the paper provide to support its claims
What specific tables and figures support the paper’s claims
Note
In your papers, I’m particularly interested in your ability to take a theoretical claim and map it onto an empirical result. Give me page numbers, tables, estimates.
What are the conclusions
What have we learned?
What do we still need to know?
What would we do differently?
What are your podcasts?
Friday
Plan for Friday:
Weaver Prowse and Piston
For Friday:
- Read Jerit and and Zhao
Revised Groups
Readings for Political Cognition
By Monday:
- Zaller and Feldman (1992)
By Wednesday:
- Lodge and Taber (2013)
NYT: Registered vs Likely Voters
More people say they did vote or will vote than actually do
Different polls use different likely voter models, relying on some combination:
- Survey questions (intent to vote, do you know where your polling place is)
- Voter files (past voting history)
Two classification approaches:
- Cutoff/Threshold (Most surveys)
- Probabilistic models (Times/Sienna Poll)
Political Knowledge
What should citizens know about politics?
Take a few moments to write down your thoughts:
What kinds of knowledge do citizens in a democracy need to possess
Do citizens generally possess this knowledge?
Is this a problem?
Luskin (1987)
- Luskin offers a definition of “political belief systems” defined in terms of
- size (number of cognitions)
- range (diversity)
- organization (constraint)
- Converse’s focus on constraint, Delli Carpini and Keeter (1993) argue “putting the cart before the dead horse.”
How Should We Define Political Knowledge
- Delli Carpini and Keeter (1996) define political knowledge as: “the range of factual information about politics that is stored in long-term memory.”
How Should We Measure Political Knowledge
- Open-ended questions can be problematic
- Instead, scholars of developed set of questions which we use to measure this concept
What do you know!
Do you happen to know what job or office is now held by Kamala Harris?
Whose responsibility is it to determine if a law is constitutional or not … is it the president, the Congress, or the Supreme Court?
How much of a majority is required for the US Senate and House to override a presidential veto?
Do you happen to know which party has the most members in the House of Representatives in Washington?
Would you say that one of the parties is more conservative than the other at the national level? Which party is more conservative?
What do you know!
Do you happen to know what job or office is now held by Kamala Harris? Vice President
Whose responsibility is it to determine if a law is constitutional or not … is it the president, the Congress, or the Supreme Court? Supreme Court
How much of a majority is required for the US Senate and House to override a presidential veto? Two-thirds
Do you happen to know which party has the most members in the House of Representatives in Washington? Republicans
Would you say that one of the parties is more conservative than the other at the national level? Which party is more conservative? Republican Party
What does the public know?
What Makes a Good Measure
- These standard measures of political knowledge have some nice properties
- They “scale” together well
- They discriminate levels of political knowledge
What Makes a Good Measure? Mix Difficulties
What Makes a Good Measure? High “Discrimination”
What Makes a Good Measure
- These standard measures of political knowledge have some nice properties
- They “scale” together well
- They discriminate levels of “political knowledge
- They predict attitudes and behavior
- But what are these scales really measuring?
- Why would we expect an increase in this measure of PK to increase voting?
Lupia (2016)
- Offers an an important distinction between:
- Information
- Knowledge
- Competence
Lupia (2016)
Information: “Information is what educators can convey to others”
Knowledge is memories of how concepts and objects are related to each other.
- Knowledge requires Information
Competence is the ability to perform a task in a particular way
Competence requires knowledge, but not complete knowledge. Depends on the context/decision
- Cues and heuristics can help
- Competence is contextual and political
What is necessary and sufficient to make competent decisions?
Political Knowledge in the 2020 ANES
Political Knowledge
Let’s take a look at political knowledge in the 2020 American National Election Study as measured by four items:
- Length of Senate Term
- Government spending on Foreign Aid
- Party control of House
- Party control of Senate
## ---- Recode data ----
df %>%
mutate(
pk_senate_term = case_when(
V201644 == 6 ~ 1,
V201644 < 0 ~ NA,
T ~ 0
),
pk_foreign_aid = case_when(
V201645 == 1 ~ 1,
V201645 < 0 ~ NA,
T ~ 0
),
pk_house = case_when(
V201646 == 1 ~ 1,
V201646 < 0 ~ NA,
T ~ 0
),
pk_senate = case_when(
V201647 == 2 ~ 1,
V201647 < 0 ~ NA,
T ~ 0
),
sex = ifelse(V201600 == 2, "Female", "Male"),
college_degree = ifelse(V201511x > 3, "College Degree", "No College"),
race = case_when(
V201549x == 1 ~ "White",
V201549x == 2 ~ "Black",
V201549x == 3 ~ "Hispanic",
V201549x == 4 ~ "Asian",
T ~ "Other"
),
race = forcats::fct_infreq(race),
income = case_when(
V201617x < 0 ~ NA,
T ~ V201617x
),
political_interest = ifelse(
V201005 < 0, NA, (V201005 - 5)*-1
)
) %>%
mutate(
political_knowledge = rowSums(
select(.,starts_with("pk")),
na.rm=T
)
)-> df## ---- Figures ----
## Individual items
df %>%
summarise(
`Senate Term` = mean(pk_senate_term,na.rm=T),
`Foreign Aid` = mean(pk_foreign_aid, na.rm = T),
`Dem House` = mean(pk_house, na.rm=T),
`Rep Senate` = mean(pk_senate, na.rm = T)
) %>%
pivot_longer(
cols = 1:4,
names_to = "Item"
) %>%
mutate(
Item = forcats::fct_reorder(Item,value),
`Percent Correct` = round(value*100,2)
) %>%
ggplot(aes(Item, `Percent Correct`))+
geom_bar(stat = "identity")+
coord_flip()+
geom_text(aes(label = scales::percent(value)),
hjust = -.25)+
ylim(0,100)+
labs(
title="Individual Knowledge Items",
x = ""
) +
theme_minimal()-> fig1
## Knowledge Scale
df %>%
group_by(political_knowledge) %>%
summarise(
n = n(),
prop = n()/nrow(df),
Percent = scales::percent(prop)
) %>%
ggplot(aes(political_knowledge, prop))+
geom_bar(stat = "identity")+
scale_y_continuous(labels = label_percent())+
geom_text(aes(y = prop, x= political_knowledge,label = Percent), vjust = -.5)+
ylim(0, .33)+
labs(
x = "Number of Items Correct",
title = "Political Knowledge Scale"
) +
theme_minimal() -> fig2
# ---- Knowledge Gaps ----
# Education
df %>%
ggplot(aes(college_degree, political_knowledge,
col = college_degree))+
stat_summary()+
guides(col = "none")+
theme_minimal()+
labs(
x= "",
y = "Average Political Knowledge",
title = "Education"
) -> fig3
# Sex
df %>%
ggplot(aes(sex, political_knowledge,
col = sex))+
stat_summary()+
guides(col = "none")+
theme_minimal()+
labs(
x= "",
y = "Average Political Knowledge",
title = "Sex"
) -> fig4
# Race
df %>%
ggplot(aes(race, political_knowledge,
col = race))+
stat_summary()+
guides(col = "none")+
theme_minimal()+
labs(
x= "",
y = "Average Political Knowledge",
title = "Race"
) -> fig5
# Income
df %>%
ggplot(aes(income, political_knowledge,
))+
stat_summary(size=.2)+
theme_minimal()+
labs(
x= "",
y = "Average Political Knowledge",
title = "Income"
) -> fig6# ---- Models ----
## Bivariate
m1 <- lm_robust(political_knowledge ~ college_degree, df)
m2 <- lm_robust(political_knowledge ~ sex, df)
m3 <- lm_robust(political_knowledge ~ race, df)
m4 <- lm_robust(political_knowledge ~ income, df)
## Multiple Regression
m5 <- lm_robust(political_knowledge ~
college_degree + sex + race + income + political_interest , df)
Knowledge Gaps
Knowledge Gaps
We could also explore knowledge gaps using regression. Recall:
Regression is tool for estimating conditional means
Regression partitions variance explained by specific factors
| Model 1 | Model 2 | Model 3 | Model 4 | Model 5 | |
|---|---|---|---|---|---|
| (Intercept) | 2.57*** | 2.00*** | 2.30*** | 1.73*** | 1.42*** |
| (0.02) | (0.02) | (0.02) | (0.03) | (0.05) | |
| No BA | -0.66*** | -0.46*** | |||
| (0.03) | (0.03) | ||||
| Male | 0.44*** | 0.34*** | |||
| (0.03) | (0.02) | ||||
| Hispanic | -0.40*** | -0.18*** | |||
| (0.05) | (0.04) | ||||
| Black | -0.52*** | -0.26*** | |||
| (0.04) | (0.04) | ||||
| Other | -0.30*** | -0.08 | |||
| (0.05) | (0.05) | ||||
| Asian | 0.12 | 0.10 | |||
| (0.07) | (0.06) | ||||
| Income | 0.04*** | 0.02*** | |||
| (0.00) | (0.00) | ||||
| Political Interest | 0.27*** | ||||
| (0.01) | |||||
| R2 | 0.07 | 0.03 | 0.02 | 0.06 | 0.19 |
| Adj. R2 | 0.07 | 0.03 | 0.02 | 0.06 | 0.19 |
| Num. obs. | 8280 | 8280 | 8280 | 7665 | 7664 |
| RMSE | 1.16 | 1.18 | 1.19 | 1.15 | 1.07 |
| ***p < 0.001; **p < 0.01; *p < 0.05 | |||||
Jerit, Barabas, and Bolsen (2006)
What’s the research question
What’s the point of this study?
Think, pair, share…
What’s the research question
- Jerit et al. argue that political knowledge is not a static concept, but instead varies over time as a function of the characteristics of individuals and their environments
What’s the theoretical framework
- What debate do the authors speak to?
- What are their theoretical contributions to this debate?
- What are the specific expectations from this theory
What’s the theoretical framework
- Jerit et al. address a literature focused on knowledge gaps which:
- highlights the importance of relatively stable individual factors like education, income, gender, race in predicting differences in knowledge
- Concludes that any changes in political knowledge are likely to benefit the “informationaly rich”
What’s the theoretical framework
- Jerit et al. lay the foundation for a distinction made by Barabas et al. (2014) between the type and temporal dimensions of knowledge
What’s the theoretical framework
- Jerit et al. argue political knowledge will vary as both a function of characteristics of the individual (education) and the information environments in which they exist
What’s the theoretical framework
From page 268:
- H1a: Increases in the overall amount of media attention to an issue will increase the average amount of knowledge in the population
- H1b: The gap in knowledge between individuals with low and high levels of education also will increase
- H2a: All else held constant, increasing the amount of newspaper coverage will raise the average level of knowledge in the population, but it should primarily benefit those with high levels of education
- H2b: An increase in television coverage will raise the average level of knowledge in the population, but it will not alter the relationship between education and knowledge
What’s the empirical design
- What data and methods do the authors use?
What’s the empirical design
- Measure specific surveillance knowledge from surveys on 41 issues (primarily health related), paired with content analysis of issue-specific coverage in the AP, Broadcast TV News, and USA Today.
- Study 1 looks at variation in coverage across two issues using a probit regression
- Study 2 looks at variation in coverage and individual education across multiple issues using multilevel modeling
😉 What you need to know (WYNK)
- “Dichotomous” taking 1 of two values (1 if correct, 0 otherwise)
- “We ran a multivariate probit…” They did a special type of regression for data that has a binary (0 or 1) outcome
- “A Multilevel Model” They pool data together across surveys and model variation at the level of both the individual i and the issue j
- “The grand mean” The mean for the entire sample
- “ANOVA” Analysis of Variance: How much variation is at the individual vs issue level
What’s are the results
- What are the hypotheses?
- What figures and tables provide evidence in support of each claim?
What are the results
- H1a
- Text page 271, Figure 2, Figure 3, Table 1, Figure 4
- Text page 271, Figure 2, Figure 3, Table 1, Figure 4
- H1b
- Figure 2, Figure 3, Table 1, Figure 4
- H2a
- Table 1, Figure 4
- H2b
- Table 1, Figure 4
Increasing Media Coverage Increases Knowledge (H1a)
Knowledge Gaps Increase When There is More Coverage (H1b)
Is this Always True?
Is this Always True?
Increasing Media Coverage Increases Knowledge (H1a)
Knowledge Gaps Increase with more Media Coverage (H1b)
Newspaper Coverage Benefits the Informationally Rich (H2a)
Newspaper Coverage Benefits the Informationally Rich (H2a)
Television Coverage Offers Wider Benefits (H2b)
Media Coverage and Knowledge Gaps
What’s are the conclusions
- What are some potential concerns about this study?
- What are the broader conclusions
Weaver, Prowse, and Piston (2019)
Take a few moments
- What’s the research question
- What’s the theoretical framework
- What’s the empirical design
- What are the results
- What are the conclusions
What’s the research question
- Do citizens have too much power given their relatively low levels of political knowledge or is it in fact the case that citizens in highly policed communities have in fact too much knowledge and not enough power
What’s the theoretical framework
WPP address a literature on political knowledge that
Finds citizens have low levels of general and factual knowledge
Worries that they may exert too much influence over politics and policy (Quirk and Hinchcliffe 1998)
What’s the theoretical framework
Following Cramer and Toff (2017) they think of political knowledge in terms of people’s lived experiences
Knowledge of facts alone is less important than how people interpret and use them that matters for civic competence
What’s the theoretical framework
WPP’s work builds on theories of policy feedback
- Policy Feedback examines the way policy experiences shape citizens’ attitudes and behaviors
WPP in particular, focus on the second face of government that deals with aspects of social control
WPP present evidence of citizen who possesses “too much” knowledge of the police, and that this knowledge is detrimental to democracy
What’s the empirical design
Conversations with Strangers
The Portal does two things simultaneously, providing the opportunity and space to discuss a topic and connecting them to those who can be assumed to be knowledgeable about that topic.
Yes, but how representative are these portals?
Portals participants are not a strictly random sample and we cannot say how representative they are of communities of interest.
We are after richer data that reveals not just a snapshot of opinion that is “representative,” but how people reason together, how they frame things in their own words not those of the survey researcher, and how they develop a theory of state action and power.
Second, we would be more concerned about representativeness or bias if we were testing hypotheses about the distributions of attitudes (how many) or causal relationships between variables (how related), studies based on a “sampling logic.
Finally, existing large-N surveys are notoriously inadequate at capturing the experiences of highly policed communities.
What are the results
Take a few momements to share with your neighbor, your understanding of the study’s findings
Two Much Knowledge, Too Little Power
The crux of their argument follows from their summary of the initial portal conversation (pp. 1155-1158) which they argue reveals:
- Knowledge gleaned from “direct experience with the police”
- Extensive factual knowledge (Names of public officials, scandals)
- Knowledge of “unwritten rules”
- Multiple sources of information (personal experience, vicarious experience, media coverage)
- Official stories vs counter narratives
What are the results
Conversations uncover
Citizens possess a “Dual Knowledge”
- What government (as represented by the Police) is supposed to do
- What government actually does
This knowledge is deep and specific (and obtained involuntarily)
Knowledge serves to “help citizens distance themselves from police oversight”
Residents of highly policed communities have “too much knowledge, too little power”
What are the conclusions
- What do we think of this paper?
- What are its strengths?
- What are its weaknesses?
- How does it change our conceptions of political knowledge
Attendance Survey
Please complete the following survey
Misinformation
What is misinformation
Kuklinski et al. (2000): “People are misinformed when they confidently hold wrong beliefs”
Related but distinct from:
- Rumors
- Conspiracy beliefs
Origins of misinformation?
Jerit and Zhao (2020) (pp 79-81) review some psychological explanations, emphasizing different cognitive motivations:
- Accuracy motives → correct decisions
- Directional motives → consistent decisions
Misinformation is form a motivated reasoning reflecting a directional desire to maintain consistency with ones’ prior beliefs.
What about the sources of misinformation?
Correcting Misinformation
Mixed and varied results
Depends on the issue, correction, and individuals
- Can you find concrete examples? (p. 83)
Possibility for corrections to backfire
Backfire Effects
Measuring misinformation
Conceptually, misinformation involves confidently holding false beliefs
- But many studies fail to measure confidence
Some scholars have proposed that misinformation is a form of expressive responding or partisan cheerleading and that partisan gaps disapper when we incentivize correct responses (Bullock et al. 2015)
- So which beliefs matter?
What do we make of the directions for further research on p. 88?
Next week
Readings for Political Cognition
By Monday:
- Zaller and Feldman (1992)
By Wednesday:
- Lodge and Taber (2013)
References
Bullock, John G, Alan S Gerber, Seth J Hill, and Gregory Huber. 2015. “Partisan bias in factual beliefs about politics.” In Quarterly Journal of Political Science, 10:519–78.
Haglin, Kathryn. 2017. “The limitations of the backfire effect.” Research & Politics 4 (3): 205316801771654.
Jerit, Jennifer, and Yangzi Zhao. 2020. “Political misinformation.” Annual Review of Political Science 23 (1): 77–94.
Kuklinski, James H, Paul J Quirk, Jennifer Jerit, David Schwieder, and Robert F Rich. 2000. “Misinformation and the Currency of Democratic Citizenship.” The Journal of Politics 62 (3): 790–816.
Lodge, Milton, and Charles S Taber. 2013. The rationalizing voter. Cambridge University Press.
Nyhan, Brendan, and Jason Reifler. 2015. “Does correcting myths about the flu vaccine work? An experimental evaluation of the effects of corrective information.” Vaccine 33 (3): 459–64.
Zaller, John, and Stanley Feldman. 1992. “A simple theory of the survey response: Answering questions versus revealing preferences.” American Political Science Review, 579–616.
POLS 1140
POLS 1140 Political Knowledge and Information Updated Mar 9, 2025