Chapter 4 Exercises: Cognitive Biases and Heuristics That Make Us Vulnerable

These exercises are designed to develop critical understanding of cognitive biases through conceptual analysis, empirical reasoning, applied case analysis, and reflective practice. Exercises span individual biases and synthesis tasks requiring integration across multiple biases.

Part A: Conceptual Foundations

Exercise 4.1 — The Gigerenzen-Kahneman Debate

Carefully consider the following two statements:

Kahneman/Tversky position: "People rely on a limited number of heuristic principles which reduce complex tasks of assessing probabilities and predicting values to simpler judgmental operations. In general, these heuristics are quite useful, but sometimes they lead to severe and systematic errors."

Gigerenzen position: "Heuristics are not inferior substitutes for optimization; rather, they exploit certain structures of the environment to enable efficient, accurate decisions."

a) What is the core empirical disagreement between these positions? Is it a disagreement about what people actually do, or about how to evaluate what they do?

b) Design an experiment that could distinguish between these two positions. What evidence would support one over the other?

c) Consider the specific context of social media news consumption. Which position better predicts the consequences of heuristic processing in this environment? Does your answer depend on what type of information is being processed?

d) Does the debate matter practically? If both agree that heuristics sometimes produce errors, does it matter whether those errors are "irrational" (Kahneman/Tversky) or "ecologically mismatched" (Gigerenzen)?

Exercise 4.2 — Availability Heuristic: Risk Perception Audit

For each of the following risks, (a) estimate the number of deaths per year in the United States from this cause without looking it up, (b) look up the actual figure from a reliable source (CDC, government statistics), and (c) calculate the ratio of your estimate to the actual figure.

• Shark attacks
• Lightning strikes
• Commercial airline crashes
• Car accidents
• Heart disease
• Medical errors (preventable deaths in hospitals)
• Gun homicides
• Terrorism (domestic)
• Falling from ladders
• Drowning

Then write a 400-word reflection addressing: 1. Which causes did you most overestimate? Which did you most underestimate? 2. How does media coverage of these causes compare to their actual death rates? 3. What does this exercise reveal about the availability heuristic in your own cognition? 4. How might these misperceptions affect public policy priorities?

Exercise 4.3 — Base Rate Neglect: The Medical Test Problem

Work through the following problem carefully, showing all reasoning:

A disease affects 1 in 1,000 people in a population. A screening test for this disease has 99% sensitivity (correctly identifies 99% of true positives) and 99% specificity (correctly identifies 99% of true negatives—i.e., has a 1% false positive rate).

You test positive. What is the probability that you actually have the disease?

a) Calculate the answer using Bayes' theorem or a frequency tree. Show your work.

b) What do most people (including many physicians) intuitively estimate? Why does the representativeness heuristic lead to systematic overestimation?

c) What is the "base rate" in this problem, and how does base rate neglect produce the error?

d) Rework the problem: if the disease affects 1 in 10 people (much more common), how does your answer change? What does this tell you about the importance of base rates?

e) How does this problem apply to the evaluation of rare events reported in news? Give a specific example involving crime statistics or health claims.

Exercise 4.4 — Anchoring: Numerical Framing Analysis

For each of the following pairs of framings of the same underlying statistic, identify the anchor that each framing provides and predict the likely direction of difference in perception between the two framings:

a) Frame A: "3,000 people died last year in workplace accidents." Frame B: "In the United States, workplace accidents kill 3,000 people per year—fewer than 10% of the annual toll from automobile accidents."

b) Frame A: "The drug doubles your risk of heart attack." Frame B: "The drug increases your absolute risk of heart attack from 0.1% to 0.2%."

c) Frame A: "56% of voters support the new immigration policy." Frame B: "44% of voters oppose the new immigration policy."

d) Frame A: "Scientists are 97% confident that human-caused climate change is real." Frame B: "3% of climate scientists do not attribute recent warming to human activity."

For each pair, discuss: (1) Which frame is more accurate? (2) Which frame is more emotionally compelling? (3) What are the implications for responsible statistical communication?

Exercise 4.5 — Confirmation Bias: The Wason Selection Task

Complete the following Wason selection task:

You are presented with four cards. Each card has a number on one side and a letter on the other. The four cards show: E, K, 4, 7

The rule to test is: "If a card has a vowel on one side, it has an even number on the other side."

Which cards do you need to turn over to test this rule? (Select all that apply and explain why.)

a) Record your initial, intuitive answer.

b) Now analyze the task more carefully using logical principles. Which cards actually need to be turned over? Which cards cannot falsify the rule regardless of what is on the other side?

c) Research shows that most people get this wrong. How does confirmation bias explain the typical error?

d) Wason found that presenting the same logical structure in a concrete, social context ("People drinking beer must be 21 or older") dramatically improves performance. Why? What does this tell us about the ecological specificity of confirmation bias?

e) Design an analogous selection task using a media literacy scenario (e.g., about the reliability of sources or the accuracy of claims).

Exercise 4.6 — Confirmation Bias in Information Search

Conduct the following experiment on yourself:

Choose a controversial topic on which you hold a prior opinion (e.g., a political question, a health controversy, a social issue). Then:

a) Conduct a 15-minute internet search using the search terms you would naturally use to learn more about the topic. Record the first 10 results you examine and their main claims.

b) Now construct search terms designed to find the strongest available arguments against your prior view. Conduct a 15-minute search using these terms. Record the first 10 results.

c) Compare the two sets of search results. How different are they? What did you learn from the second search that you would not have found in the first?

d) Write a 500-word reflection on what this exercise revealed about your own confirmation bias in information search behavior.

e) What practical habit could you adopt to reduce confirmation bias in your regular information consumption?

Part B: Specific Bias Analysis

Exercise 4.7 — The Backfire Effect Debate: Evidence Evaluation

Read the following summaries of two studies and answer the questions below:

Study 1 (Nyhan & Reifler, 2010): When Republican-identified participants received a correction stating that Iraq had no WMDs before the 2003 invasion, some participants—particularly those with the strongest prior belief that Iraq had WMDs—showed increased belief in WMDs after reading the correction. The researchers interpreted this as a "backfire effect."

Study 2 (Wood & Porter, 2019): In 52 experimental conditions testing political misperceptions, corrections consistently reduced false belief regardless of partisan identity. The researchers found no evidence for backfire effects in any condition.

a) What are three methodological factors that might explain why Study 1 found backfire effects but Study 2 did not?

b) What do we lose and gain by moving from a study of specific political events (Nyhan & Reifler) to a broad study of 52 misperceptions (Wood & Porter)?

c) If corrections consistently reduce false belief (as Wood & Porter find) but the effects are modest (e.g., reducing false belief by 10-15 percentage points), what does this imply for the practical significance of fact-checking as a strategy for misinformation reduction?

d) A fact-checker asks: "Should I worry about backfire effects when correcting political misinformation?" Based on the current evidence, what is your answer?

Exercise 4.8 — Dunning-Kruger: Calibration Exercise

The following exercise is designed to measure your calibration—the alignment between your confidence and your accuracy.

For each of the following questions, provide: - Your best answer - A confidence interval: a lower bound and upper bound such that you are 90% confident the true answer falls within the range

(This is not asking what you know—it's asking you to set bounds that you believe are 90% likely to contain the true answer.)

1. What year was the first commercially available smartphone released?
2. How many countries are members of the United Nations?
3. What is the population of Brazil (to the nearest 10 million)?
4. In what year did the Berlin Wall fall?
5. How many bones are in the adult human body?
6. What percentage of the Earth's surface is covered by water?
7. In what year was the Eiffel Tower built?
8. What is the approximate distance from Earth to the Moon (in miles)?
9. How many US Supreme Court justices have been women?
10. What is the approximate population of New York City?

After completing all 10, look up the answers. Count how many true answers fell within your 90% confidence intervals. A well-calibrated person should have 9/10 (90%) of their intervals contain the true answer. Fewer means you were overconfident (intervals too narrow); more means you were underconfident.

Write a 300-word reflection on what this exercise revealed about your own calibration.

Exercise 4.9 — Cultural Cognition: Identity and Factual Belief

Kahan's cultural cognition research shows that beliefs about contested empirical questions track cultural identity dimensions. Answer the following:

a) List five empirical questions (questions with factually correct answers, not merely questions of values) where you believe individual beliefs in the United States are significantly correlated with political identity. For each, state what you believe the scientific evidence shows.

b) For one of these questions, explain in detail how identity-protective cognition could lead a person with a specific cultural identity to sincerely hold a false belief about the empirical evidence.

c) How does Kahan's finding that analytical sophistication amplifies identity-protective cognition differ from the naive prediction that more education produces better calibration?

d) If you were designing a science communication strategy to reduce identity-protective cognition on a specific contested topic, what elements would you include? Draw specifically on Kahan's framework.

e) Are there any contested empirical questions where you suspect your own beliefs might be influenced by identity-protective cognition? How would you test this?

Exercise 4.10 — Proportionality Bias and Conspiracy Thinking

For each of the following major events, (a) identify the official or scientific explanation, (b) identify a prominent conspiracy theory, (c) analyze how proportionality bias contributes to the appeal of the conspiracy theory, and (d) evaluate what additional cognitive factors might reinforce it.

1. The assassination of John F. Kennedy (1963)
2. The September 11, 2001 terrorist attacks
3. The COVID-19 pandemic origin
4. Princess Diana's death in a car accident (1997)
5. The moon landing (1969)

For one of these events, write a 500-word analysis that explains why the conspiracy theory is epistemically less justified than the mainstream explanation, while acknowledging what psychological needs it fulfills.

Part C: Applied Analysis

Exercise 4.11 — Bias Interaction Analysis

A news story reports: "New study finds immigrants are responsible for a dramatic rise in violent crime in border cities." The story includes an interview with a local resident who says "I knew it—I've seen it with my own eyes. These people are dangerous."

Identify and explain how at least five distinct cognitive biases from this chapter could contribute to a reader's acceptance of this claim. For each bias, explain: - What specific aspect of the claim or framing it would affect - In what direction it would push belief - Whether it would interact with other biases

Then: The next week, a comprehensive study finds that violent crime in these cities has actually fallen and that immigrants commit violent crimes at lower rates than native-born citizens. Identify three cognitive biases that would likely reduce the impact of this correction.

Exercise 4.12 — Social Media Platform Audit

Choose a social media platform you use regularly. Spend 30 minutes scrolling through your feed and recording:

a) Five instances where the availability heuristic may have been at work (dramatic events, rare occurrences presented as common, etc.)

b) Three instances where content seemed designed to activate in-group/out-group dynamics

c) Two instances where anchoring effects might influence how statistics are understood

d) One example of content that exploits proportionality bias (implying a major event must have a major, organized cause)

Write a 600-word analysis connecting your observations to the cognitive science reviewed in this chapter.

Exercise 4.13 — Debiasing Strategy Design

Design a debiasing intervention for one of the following contexts. Your design must: - Target at least two specific cognitive biases from this chapter - Be based on at least one debiasing strategy with empirical support - Include a realistic implementation plan - Include a plan to evaluate effectiveness - Address potential unintended consequences

Choose one context: a) A high school social studies classroom addressing political news literacy b) A corporate communications department trying to improve employees' critical evaluation of internal research claims c) A public health department trying to improve community decision-making about vaccine uptake d) A newsroom trying to reduce confirmation bias in journalists' story selection

Exercise 4.14 — The Conjunction Fallacy in Misinformation

The conjunction fallacy predicts that adding specific, plausible details to a claim makes it seem more probable even though it cannot be.

a) Identify a real piece of misinformation that exploits the conjunction fallacy by adding specific, plausible-seeming details that make an implausible core claim seem more credible. Analyze how the conjunction fallacy is being exploited.

b) Rewrite the claim without the specific, conjunction-exploiting details. Does the core claim seem more or less credible in this stripped-down form? Why?

c) What does this exercise suggest about how fact-checkers should analyze detailed conspiracy theories? Should they evaluate the core claim, the conjunction of all details, or individual details separately?

Exercise 4.15 — Calibration Improvement Plan

Based on Tetlock's research on superforecasters and the calibration exercise in Exercise 4.8:

a) Research the five specific habits of well-calibrated superforecasters described in Tetlock & Gardner (2015). List them and explain each.

b) Choose the two habits you believe would be most valuable for news literacy and explain why.

c) Design a 30-day personal practice plan for developing better calibration in your own information evaluation. Be specific about what you will do, how often, and how you will track progress.

d) What domains are likely to be hardest for you to improve calibration in, and why?

Part D: Synthesis and Reflection

Exercise 4.16 — Bias Catalog Construction

Create a personal "bias catalog" — a reference document that includes:

For each of the 10 biases covered in this chapter: - Name and one-sentence definition - The cognitive mechanism that produces it - One real-world example from news or social media - One debiasing strategy that addresses it - A personal reflection on how this bias has affected your own thinking

Exercise 4.17 — Research Brief: Availability Cascade Analysis

Select a specific risk event from the past five years that received extensive media coverage disproportionate to its statistical frequency (e.g., a particular crime wave, a disease outbreak, a product safety concern). Write a 700-word research brief that:

a) Documents the scale of media coverage using quantitative evidence (number of stories, social media shares, etc.)

b) Documents the actual base rate and risk level using government or peer-reviewed sources

c) Analyzes the availability cascade that developed, using Sunstein's framework

d) Evaluates the policy consequences of the distorted availability (what policy responses were shaped by the perception rather than the reality?)

e) Proposes what responsible media coverage would have looked like

Exercise 4.18 — Comparative Bias Analysis: Climate vs. Vaccination

Both climate change and vaccine safety have been the subject of persistent public misperception despite strong scientific consensus. Yet the misperception patterns are somewhat different: on climate change, party identity strongly predicts belief; on vaccines, the pattern is more complex.

a) Using the concepts from this chapter, explain why identity-protective cognition would predict stronger partisan divides on climate change than on vaccination.

b) On which topic is the availability heuristic more likely to drive false belief? Availability of what type of information?

c) On which topic is proportionality bias more likely to contribute to conspiracy thinking?

d) What does the different pattern of misperception on these two topics suggest about the relative importance of different cognitive mechanisms for different types of scientific consensus?

Exercise 4.19 — The Backfire Effect and Social Media Design

Twitter (now X) has experimented with prompting users who try to share articles without reading them ("You are about to share this article—do you want to read it first?"). This is a form of accuracy nudge.

a) Which cognitive biases does this intervention target?

b) Using the evidence from this chapter and Chapter 3, predict the size and direction of the effect on sharing accuracy.

c) Are there conditions under which this intervention might backfire? If so, what are they?

d) Design a more comprehensive suite of platform interventions that would address five different cognitive biases from this chapter. For each, explain the mechanism by which it reduces bias impact.

Exercise 4.20 — Anchoring in News Statistics: Case Study

Find three news articles from the past month that contain numerical statistics. For each article:

a) Identify the first statistic mentioned and explain how it functions as an anchor for subsequent interpretation

b) Identify whether the statistic is presented in absolute terms, relative terms, or both

c) Identify whether a relevant base rate is provided

d) Rewrite the opening sentence to present the same information with less biasing anchoring effects while maintaining accuracy

e) Assess: Does your rewrite feel less alarming than the original? If so, what does this say about the relationship between alarm and accuracy in news reporting?

Exercise 4.21 — In-Group/Out-Group Dynamics: Source Credibility Experiment

Design a simple survey experiment that tests whether the same statistical claim about a politically sensitive topic is evaluated differently depending on the perceived political identity of the source. Your design should include:

a) The specific claim you will test (choose a factual claim about a politically contested topic)

b) At least three source conditions varying in perceived political identity

c) At least two dependent measures (e.g., belief in the claim, credibility of the source, likelihood of sharing)

d) A control condition

e) The sample you would recruit and why

f) What results would support the cultural cognition hypothesis? What results would challenge it?

Exercise 4.22 — Debiasing the News Environment: Policy Brief

You have been commissioned to write a policy brief for a national media literacy task force. The brief should address: "What structural changes to the news and social media environment would most effectively reduce the impact of cognitive biases on public understanding of important issues?"

Your brief (600-800 words) should: - Identify three structural features of the current news environment that amplify specific cognitive biases - Propose three feasible structural changes, each addressing at least one cognitive bias - Provide evidence that each proposed change is likely to be effective - Identify potential objections to each proposal and respond to them - Conclude with a ranking of the three proposals by cost-effectiveness

Exercise 4.23 — Cognitive Bias Interaction: Case Study

The anti-vaccination movement exemplifies how multiple cognitive biases interact to produce and maintain false beliefs. Write a 800-1000 word essay analyzing how the following biases interact in the context of vaccine hesitancy:

• Availability heuristic (vaccine injury stories vs. disease statistics)
• Representativeness heuristic (temporal coincidence of vaccination and illness onset)
• Confirmation bias (selective search for adverse event reports)
• In-group/out-group bias (trust in peer communities over medical authorities)
• Dunning-Kruger effect (overconfidence in personal research)
• Proportionality bias (vaccine injuries feel too minor to explain autism—but autism also feels too major to be coincidental)

Your essay should explain how each bias contributes, how they reinforce each other, and what this analysis implies for vaccine communication strategy.

Exercise 4.24 — Tribal Epistemics: The Messenger Effect

Research on tribal epistemics suggests that the source of information matters as much as its content for identity-motivated reasoners. Evaluate the following scenarios:

a) A climate scientist presents identical climate data to two audiences: one that identifies as politically conservative, one as politically liberal. Predict the likely reception in each audience and explain using cultural cognition theory.

b) A conservative Republican politician gives a speech supporting climate action citing the same scientific evidence. How would this change the reception among conservative audiences? Would it change liberal reception?

c) A trusted community leader in a vaccine-hesitant community delivers a vaccine recommendation using the same scientific evidence as public health officials. Why might this be more effective?

d) Based on these scenarios, what does the research on tribal epistemics recommend for the design of public health communication campaigns on contested topics?

Exercise 4.25 — Synthesis Essay

Write a 1,000-1,200 word essay responding to the following prompt:

"A policymaker argues: 'The solution to misinformation is simple: educate people about cognitive biases so they can recognize them and correct for them. Once people know about the availability heuristic, they won't be misled by dramatic but rare events; once they know about confirmation bias, they'll seek out disconfirming evidence; once they know about the backfire effect, corrections will work better.' Evaluate this argument. What does the empirical evidence from this chapter tell us about the prospects for this education-based solution? What is it likely to achieve, and what are its limits?"

Your essay should: - Accurately represent the empirical evidence for and against education-based debiasing - Specifically address the "bias blind spot" evidence - Distinguish between different types of education (generic awareness, specific technique training, structural interventions) - Propose a more comprehensive debiasing framework that acknowledges the limits of pure education - Draw a defensible conclusion about the role of education in a comprehensive response to misinformation

Exercise 4.26 — Cross-Chapter Synthesis

Connect the material from Chapter 4 to concepts from Chapter 3 by completing the following mapping table:

For each Chapter 3 concept, identify which Chapter 4 bias(es) it interacts with and describe the interaction:

Complete the table with substantive entries and then write a 300-word synthesis explaining which Chapter 3-Chapter 4 interaction you believe is most consequential for real-world misinformation susceptibility and why.

Exercise 4.27 — The Adaptive Function of Bias

Cognitive biases are often discussed as defects or failures. But each bias in this chapter evolved because it served adaptive functions in ancestral environments.

For each of the following biases, explain: (a) What adaptive problem it likely evolved to solve (b) The type of environment in which it works well (c) The specific features of the modern information environment that make it fail

• Availability heuristic
• Representativeness heuristic
• Confirmation bias
• In-group favoritism
• Proportionality bias

Then write a 200-word conclusion addressing: Does understanding the adaptive origins of cognitive biases change how we should respond to them? Does it make more ethical sense to try to change the biases themselves, or to change the environment in which they operate?

Exercise 4.28 — Final Reflection: Personal Bias Portfolio

This exercise asks you to reflect on the full content of Chapter 4 and identify your own areas of greatest cognitive vulnerability.

Based on what you have learned, write a 600-800 word personal reflection addressing:

a) Which two cognitive biases from this chapter do you believe most significantly affect your own information processing? Provide specific examples from your own experience.

b) What features of your social identity, information consumption habits, and knowledge domain most amplify your vulnerability to these biases?

c) What specific, practical changes to your information consumption and evaluation practices will you implement based on this chapter?

d) What are the limits of individual cognitive improvement as a solution to the misinformation problem? What systemic changes would address the root causes rather than placing the entire burden on individual cognition?