Detailed slide-by-slide text content extracted from this presentation.
Slide 01
Cognitive Biases
- Psychology
- The systematic glitches in human reasoning that shape every decision, judgment, and belief — often without our awareness.
- A survey of the science of human irrationality · 30 slides
- 1 / 30
Slide 02
The Rational Animal — A Myth
- Introduction
- Humans are not the rational actors of classical economics. We are decision-making machines built by evolution for a world very different from the one we inhabit — and our shortcuts show.
- A cognitive bias is a systematic pattern of deviation from rationality in judgment. Unlike random errors, biases are predictable — they consistently push our thinking in the same wrong direction. They emerge from mental shortcuts (heuristics) that evolved because they were "good enough" for our ancestors' needs, even when they fail in the modern world.
- Heuristics vs. Biases
- Heuristics are mental shortcuts — fast, efficient approximations. They often work. Biases are what happens when heuristics misfire. Daniel Kahneman's "System 1" (fast, intuitive) vs. "System 2" (slow, deliberate) thinking captures this duality.
- Why We Have Them
- Evolution optimized us for speed, not accuracy. In the ancestral environment, a fast wrong judgment was often safer than a slow correct one. We carry these evolved shortcuts into a world of financial markets, medical decisions, and social media — with predictable problems.
- 2 / 30
Slide 03
The Men Who Mapped Irrationality
- History of the Field
- The modern science of cognitive bias was born in a series of landmark papers by Israeli psychologists Daniel Kahneman and Amos Tversky, beginning in 1974. Their collaboration produced the most influential body of work in the history of psychology.
- The Partnership
- Kahneman and Tversky worked together for over a decade, each deeply complementing the other's strengths. Tversky (the mathematical genius) and Kahneman (the intuitive psychologist) produced work neither could have done alone. Their papers — "Judgment Under Uncertainty" (1974), "Prospect Theory" (1979) — redefined behavioral science.
- The Nobel Prize
- In 2002, Kahneman was awarded the Nobel Prize in Economics — remarkable for a psychologist. (Tversky had died in 1996; Nobel Prizes are not awarded posthumously.) The prize recognized that human economic behavior is systematically irrational in predictable ways — undermining classical economics' rational actor model.
- Key Text
- Kahneman's Thinking, Fast and Slow (2011) remains the definitive popular account of cognitive bias research — one of the most cited popular science books ever written.
- 3 / 30
Slide 04
Confirmation Bias
- Social & Information Biases
- The tendency to search for, interpret, favor, and recall information in a way that confirms one's preexisting beliefs — arguably the most consequential bias in public life.
- "What the human mind wishes to be true, it readily believes."
- — Francis Bacon, Novum Organum, 1620
- How It Works
- We expose ourselves to confirming information, interpret ambiguous evidence as supporting our view, and remember confirming facts more readily than contradicting ones. Crucially, we ask different questions of information we distrust — looking for any flaw — versus information we want to believe.
- Real-World Impact
- Social media algorithms amplify it — echo chambers
- Medical misdiagnosis when doctors anchor on first hypothesis
- Political polarization — each side sees confirming evidence
- Investment bubbles — investors ignore contradicting data
- Scientific replication crisis — researchers see desired results
- 4 / 30
Slide 05
The Availability Heuristic
- Probability & Risk
- We judge the probability of events by how easily examples come to mind. Vivid, recent, or emotionally charged events feel more probable — regardless of their actual frequency.
- Classic Examples
- People overestimate death by airplane and shark attack; underestimate death by car and heart disease
- After watching a crime documentary, people overestimate local crime rates
- Stock market investors overweight recent crashes in their risk models
- Doctors who just saw a rare condition over-diagnose it in subsequent patients
- The Media Mechanism
- News organizations select for dramatic, unusual events — making rare occurrences seem common. The "mean world syndrome" — the belief that the world is more dangerous than statistics show — is a product of media availability effects on risk perception.
- Kahneman's Demonstration
- Are there more words in English that start with the letter K, or have K as the third letter?
- Most people guess K-initial. In reality, three times as many words have K as the third letter — but K-initial words come to mind more easily (kitchen, king, key...). Availability ≠ frequency.
- Correcting the Bias
- Statistical literacy helps — learning to compare perceived risk against base rates. Ask: "Is this coming to mind because it's common, or because it's memorable?"
- 5 / 30
Slide 06
Anchoring Bias
- Judgment & Estimation
- The first number we encounter sets an "anchor" that disproportionately influences all subsequent estimates — even when the anchor is arbitrary or irrelevant.
- The Discovery
- Kahneman and Tversky's original experiment: a wheel of fortune was spun (rigged to land on 10 or 65). Subjects then estimated what percentage of African countries were in the UN. Those who saw 65 guessed higher than those who saw 10 — a random number had corrupted their judgment.
- In Negotiations
- Whoever makes the first offer in a negotiation sets the anchor. High initial asking prices for houses result in higher final sale prices. Legal damages are higher when prosecutors request higher sums. Salary negotiations: first number stated tends to dominate the final outcome.
- In Commerce
- "Was $199, now $99" — the $199 anchors value perception
- Restaurant wine lists put expensive bottles first to make $60 bottles seem reasonable
- Salary anchoring: always stating a high number first
- Charitable donations increase when a suggested amount is provided — even if arbitrary
- Key Insight
- Anchoring persists even when people know about it, even when the anchor is explicitly random, and even when experts are warned. It is among the most robust and resistant-to-correction of all biases.
- 6 / 30
Slide 07
Loss Aversion
- Prospect Theory
- Losses loom larger than gains — psychologically, losing $100 feels roughly twice as bad as gaining $100 feels good. This asymmetry distorts rational decision-making in profound ways.
- Losses are approximately twice as powerful as gains in their psychological impact.
- Prospect Theory
- Kahneman and Tversky's 1979 paper "Prospect Theory" showed that people evaluate outcomes relative to a reference point, and that the pain of loss is disproportionate to the pleasure of equivalent gain. This is one of the most replicated findings in behavioral science.
- The Endowment Effect
- People value things they own more than identical things they don't own — merely because ownership triggers loss aversion. In experiments, people demanded significantly more to give up an object they'd been given than they were willing to pay to acquire the same object.
- Consequences
- Investors hold losing stocks too long (hoping to "break even")
- Gamblers chase losses to recover ("just one more bet")
- Political messaging: "Don't lose X" outperforms "Gain X"
- Insurance over-purchase relative to rational calculations
- Risk aversion in gains, risk-seeking in losses — the disposition effect
- Workers resist pay cuts more than they value equivalent raises
- 7 / 30
Slide 08
The Dunning-Kruger Effect
- Metacognition
- People with limited knowledge in a domain tend to overestimate their competence — while genuine experts often underestimate theirs. Incompetence prevents self-recognition of incompetence.
- The 1999 Study
- David Dunning and Justin Kruger had Cornell undergraduates take tests of humor, logic, and grammar, then estimate their performance relative to peers. The bottom quartile dramatically overestimated their performance — they lacked the metacognitive skill to recognize their errors. The top quartile, meanwhile, assumed others performed as well as they did.
- The Mechanism
- The skills needed to perform well in a domain are largely the same skills needed to recognize competent performance. As Bertrand Russell put it: "The fundamental cause of the trouble is that in the modern world the stupid are cocksure while the intelligent are full of doubt."
- The Learning Curve
- Competence tends to follow a characteristic arc:
- Beginner: false confidence ("how hard can it be?")
- Initial exposure: crushing realization of ignorance
- Intermediate: gradual rebuilding of calibrated confidence
- Expert: accurate self-assessment, often with persistent humility
- Mitigations
- Exposure to expertise, deliberate practice, and feedback from rigorous sources all help calibrate self-assessment. The Socratic insight — knowing what you don't know — is the beginning of intellectual growth.
- 8 / 30
Slide 09
The Halo Effect
- Social Perception
- A positive impression in one area creates a "halo" that influences perceptions in unrelated areas. Physical attractiveness, status, or a single impressive achievement color our entire judgment of a person.
- Edward Thorndike's Discovery
- Psychologist Edward Thorndike coined the term in 1920 after observing that military officers' ratings of soldiers on one trait (e.g., intelligence) correlated strongly with ratings on apparently unrelated traits (e.g., physical bearing) — suggesting a general positive or negative impression was driving all ratings.
- Physical Attractiveness
- Beautiful people are assumed to be more intelligent, more competent, more moral, and more successful. They receive lighter criminal sentences, higher salaries, and more favorable evaluations. This is the attractiveness halo — one of the most thoroughly documented biases in social psychology.
- The Business Halo
- Companies doing well financially are perceived as having better strategy, leadership, and culture — regardless of actual evidence. When performance declines, the same attributes are suddenly viewed negatively. Phil Rosenzweig's "The Halo Effect" (2007) documents how management bestsellers systematically confuse correlation with causation through halo-contaminated data.
- Election Science
- Studies show taller candidates win presidential elections more often than expected. In multiple studies, research participants shown brief video clips of candidates (with audio muted) predicted election outcomes better than chance — judging on perceived competence signaled by appearance alone.
- 9 / 30
Slide 10
The Sunk Cost Fallacy
- Decision-Making
- We continue investing in failing endeavors because of past investments — money, time, or effort already spent — rather than evaluating future prospects on their merits. Economists call these "sunk costs" because they cannot be recovered.
- "Rational decision-making requires ignoring sunk costs. But humans almost never do."
- Classic Examples
- Staying in a bad relationship because of "how much time we've invested"
- Watching a boring film through to the end because you paid for it
- Continuing a failing military campaign to justify earlier sacrifices
- The Concorde fallacy: continuing to fund an economically doomed project
- Investors holding losing stocks rather than cutting losses
- The Concorde Fallacy
- The British and French governments continued funding the Concorde supersonic jet long after it was clear it would never be commercially viable — unwilling to acknowledge that the already-spent millions were gone regardless of future decisions. The project continued for political and psychological reasons that had nothing to do with rational assessment of future value.
- The Remedy
- Ask only: "If I were starting fresh today, would I choose this?" Past investment is irrelevant to the correct future decision. This is much easier to say than to do.
- 10 / 30
Slide 11
Hindsight Bias
- Memory & Judgment
- After an event occurs, people believe they "knew it all along" — overestimating the predictability of past events. History looks inevitable in retrospect; at the time, it rarely was.
- The Mechanism
- When we learn an outcome, our memory reconstructs our prior beliefs to be more consistent with what actually happened. We literally remember ourselves as having been wiser than we were. This makes learning from experience difficult — if we think we knew it all along, we don't update our models.
- In Medicine
- Medical case review studies consistently show that evaluators who are told a patient died judge the initial physicians as more negligent than evaluators who don't know the outcome — even when reading identical case notes. Hindsight makes bad outcomes seem predictable that were not.
- The 9/11 Example
- Intelligence professionals who failed to connect the dots before 9/11 were widely criticized as negligent. But studies showed that the same warnings, presented without knowledge of the outcome, were judged by other professionals as reasonable not to have acted upon. Hindsight made the failure seem obvious — at the time, it was not.
- Philip Tetlock's Research
- Tetlock's work on political forecasting showed that experts' predictions were barely better than chance — yet those same experts consistently remembered having been right. Hindsight bias protected their self-image as forecasters while preventing the learning that would actually improve prediction.
- 11 / 30
Slide 12
The Fundamental Attribution Error
- Social Psychology
- When explaining others' behavior, we overemphasize their character or disposition and underestimate situational factors. When explaining our own behavior, we do the reverse. It is arguably the most documented error in social psychology.
- The Asymmetry
- When you cut me off in traffic: you're a reckless, selfish person. When I cut you off: I was late for something important, the lane was unclear, you weren't signaling. Same behavior — radically different explanations depending on whose perspective we take.
- Lee Ross's Research
- Psychologist Lee Ross (who named the error in 1977) demonstrated that people consistently overattribute behavior to personality even when shown that behavior was situationally determined — even when they were told the person had been randomly assigned to a role or position.
- Consequences
- Blaming poverty on personal failings, ignoring structural causes
- Assuming criminals are fundamentally bad people, not products of circumstances
- Under-sympathizing with people in difficult situations
- Misunderstanding why people hold different political views
- The Actor-Observer Difference
- This bias also explains why we judge ourselves more leniently than we judge others — we have full access to our own situational constraints, but only see others' behavior from the outside, making dispositional explanations natural.
- 12 / 30
Slide 13
Status Quo Bias
- Change & Inertia
- We prefer the current state of affairs to any change — even when change would be objectively beneficial. The status quo functions as the reference point from which all alternatives are perceived as losses.
- Organ Donation Rates
- Countries where organ donation is the default (opt-out) have donation rates over 90%. Countries requiring opt-in have rates below 20%. Same population, vastly different outcomes — purely from status quo framing.
- Financial Inertia
- Most employees stick with default retirement fund allocations even when alternatives would be more appropriate. The default is the status quo — changing it requires active effort that most people never make.
- Policy Change
- Reform is systematically harder than the reverse. People's resistance to change is not proportional to the magnitude of the change but to its perceived deviation from the current state.
- Richard Thaler's Insight
- "Nudge" theory exploits status quo bias constructively: by making the desirable option the default, governments and organizations can dramatically shift behavior without restricting choice. The insight won Thaler the 2017 Nobel Prize in Economics.
- 13 / 30
Slide 14
The Planning Fallacy
- Time & Effort Estimation
- We systematically underestimate the time, costs, and risks of future projects while overestimating their benefits — even when we know that similar past projects have overrun.
- The Optimism Gap
- The planning fallacy was named by Kahneman and Tversky in 1979. It applies at every scale: students underestimate essay completion time, contractors underestimate renovation costs, governments underestimate infrastructure budgets. The Sydney Opera House was projected at $7 million; it cost $102 million and was delivered ten years late.
- Why We Do It
- We focus on the specific plan rather than the "reference class" of similar projects. We imagine things going well — the "inside view." We fail to account for unknown unknowns. We are also motivated to be optimistic: honest plans don't get funded.
- Major Project Overruns
- Denver International Airport: 16 months late, 200% over budget
- Boston's "Big Dig": projected $2.8 billion; actual cost $14.6 billion
- Edinburgh trams: £375 million projected; £776 million actual
- Berlin Brandenburg Airport: opened 9 years late; 3× over budget
- The Remedy: Reference Class Forecasting
- Kahneman's solution: before making your optimistic estimate, look at how similar projects actually turned out. Take the "outside view" — the base rate of similar endeavors. This simple step dramatically improves forecast accuracy.
- 14 / 30
Slide 15
In-Group Favoritism
- Social Identity
- We favor members of our own group over outsiders — attributing more positive traits to in-group members, making more charitable interpretations of their behavior, and showing higher degrees of trust and cooperation.
- Minimal Group Paradigm
- Henri Tajfel's landmark 1970 experiments showed that people display in-group favoritism even when divided into groups by entirely arbitrary criteria — preference for Kandinsky vs. Klee paintings, coin tosses, or random assignment. The mere act of being categorized creates discrimination. No history, no conflict, no real difference needed.
- Social Identity Theory
- Tajfel and Turner's Social Identity Theory explains the mechanism: our sense of self-worth is partly derived from our group membership. When our group compares favorably to others, our self-esteem benefits. We are therefore motivated to see our group positively — and rival groups negatively.
- Real-World Manifestations
- Racial and ethnic discrimination
- Nationalist favoritism in international policy
- Sports fan irrationality ("my team's victory is more deserved")
- Workplace nepotism and same-university hiring preferences
- Jury bias based on defendant's perceived social group
- Political polarization ("the other side is not just wrong, but evil")
- The Contact Hypothesis
- Gordon Allport's finding: extended, equal-status contact between groups under cooperative conditions reduces prejudice. The key is "cooperative" — competition intensifies bias. Shared goals dissolve group boundaries.
- 15 / 30
Slide 16
The Framing Effect
- Language & Decision
- People respond differently to the same information depending on how it is presented. A choice framed as a gain leads to different decisions than the same choice framed as a loss — even when the outcomes are mathematically identical.
- The Classic Experiment
- Kahneman and Tversky's "Asian Disease Problem": participants were told 600 people faced a disease. Program A would save 200 lives. Program B had a 1/3 probability of saving all 600, and 2/3 of saving no one. Most chose A (the certain option). When the same programs were reframed — Program A: 400 people will die; Program B: 1/3 chance no one dies — most switched to B. The probabilities were identical; the framing was not.
- Consequences
- Medical consent: "90% survival rate" vs. "10% mortality rate" produces different patient choices
- Food labeling: "95% fat-free" vs. "5% fat" — same product, different perception
- Political messaging: "Death tax" vs. "Estate tax" — same policy, different support levels
- Negotiation: "We're 10% apart" vs. "We agree on 90%" — different impasse rates
- Implication
- There is no "neutral" frame. All communication involves framing choices. The illusion of purely factual, objective presentation is itself a frame.
- 16 / 30
Slide 17
Survivorship Bias
- Statistics & Sampling
- We focus on those who have "survived" some selection process and ignore those who did not — leading to systematically wrong conclusions about what produced success.
- The WWII Example
- During WWII, the U.S. military asked statistician Abraham Wald to help armor bombers. They showed him damage maps of returning aircraft — bullet holes concentrated in the fuselage and wings. The obvious conclusion: armor those areas. Wald's insight: look where the returning planes were NOT hit. Planes that were hit in the engines didn't return. The data only showed survivors. Wald recommended armoring engines — and was right.
- Modern Examples
- Business success: We study successful companies, ignoring the far more numerous failures — drawing false lessons from an unrepresentative sample
- "School of hard knocks": Successful people who lacked formal education; we don't see those who also lacked education and failed
- Investment funds: Failed funds close; performance databases only contain surviving funds — past returns look better than they were
- Old buildings: We admire durable old buildings; the many that collapsed aren't visible to study
- 17 / 30
Slide 18
Overconfidence
- Self-Assessment
- People consistently overestimate their own abilities, the accuracy of their knowledge, and the precision of their forecasts. It is one of the most robust and consequential biases in the literature.
- 93%
- of U.S. drivers who rate themselves above average — mathematically impossible
- 68%
- of lawyers predict victory before going to trial — can't all be right
- 74%
- of fund managers who believe they outperform the market — fewer than 25% actually do
- Calibration Failure
- A well-calibrated person who says "I'm 90% confident" should be right 90% of the time. In practice, people stating 90% confidence are typically right around 70–75% of the time — a significant calibration gap. Experts in many fields show calibration no better than novices.
- Why It Persists
- Overconfidence is adaptive in some contexts — confidence signals status, attracts mates, wins followers. People prefer confident advisors even when calibrated uncertainty would serve them better. The social rewards of confidence keep overconfidence evolutionarily stable even when epistemically costly.
- 18 / 30
Slide 19
The Representativeness Heuristic
- Pattern Recognition
- We judge the probability of events by how much they match our mental prototypes of categories — ignoring base rates, sample sizes, and statistical regression.
- The Linda Problem
- Kahneman and Tversky's classic: "Linda is 31 years old, single, outspoken, very bright. She majored in philosophy. She was deeply concerned with issues of discrimination and social justice, and participated in anti-nuclear demonstrations. Which is more probable: A) Linda is a bank teller; B) Linda is a bank teller who is active in the feminist movement?"
- 85% of subjects chose B — even though A is logically more probable. Adding "feminist movement" made Linda more representative of our mental image — but statistically less probable. This is the "conjunction fallacy."
- Base Rate Neglect
- When someone is described as "shy, loves books, detail-oriented" and asked whether they're a librarian or a farmer — most say librarian. But there are many more farmers than librarians; even if more librarians fit this description, the base rates dominate. We ignore the larger category in favor of representativeness.
- Gambler's Fallacy
- A coin lands heads five times in a row. Most people believe tails is "due." But each flip is independent — past results don't change future probabilities. The gambler's fallacy is representativeness applied to random sequences: we expect short sequences to "look" random even when they needn't.
- 19 / 30
Slide 20
Priming
- Unconscious Influence
- Exposure to one stimulus unconsciously influences response to later stimuli — without awareness of the connection. Our thinking is more contextually influenced than we believe.
- How It Works
- Reading words associated with old age (Florida, wrinkles, bingo) caused participants in a famous study to walk more slowly afterward — they had been primed with concepts of aging. Financial decisions are affected by whether there's a briefcase or a backpack visible in the room. Judges give harsher sentences in the hour before lunch (the "hungry judge effect").
- The Replication Problem
- Some of the most dramatic priming findings — including the old-age walking study — have failed to replicate reliably. This is part of the broader "replication crisis" in social psychology. The existence of priming is not in doubt; the magnitude of some specific effects is. Scientists are actively recalibrating.
- Reliable Priming Effects
- Semantic priming (doctor primes nurse — faster recognition) — very robust
- Mood priming — affect colors perception and recall
- Hunger affecting judgment — replicated across cultures
- Hot vs. cold beverages affecting warmth judgments of strangers
- Background music affecting wine selection in stores
- Bottom Line
- Context shapes thinking more than we intuitively believe. The fantasy of purely context-independent rational thought is itself a bias — the bias of believing we're unbiased.
- 20 / 30
Slide 21
Choice Architecture and Nudges
- Behavioral Design
- If cognitive biases are predictable and systematic, they can be exploited — or corrected — through thoughtful design of the decision environment. This is the insight of "nudge" theory.
- Richard Thaler and Cass Sunstein
- In their 2008 book Nudge, behavioral economist Thaler and legal scholar Sunstein argued that policymakers should design "choice architectures" that leverage cognitive biases toward better outcomes — without restricting freedom. The U.K. established a "Behavioural Insights Team" (the "Nudge Unit") to apply these principles to government policy.
- The SAVE More Tomorrow Program
- Thaler and Shlomo Benartzi created a retirement savings program where employees committed future pay raises to savings — leveraging loss aversion (they didn't lose current income), present bias (the loss was future), and inertia (automatic enrollment). Participation rates tripled. It is now used by thousands of U.S. employers.
- Nudge Examples
- Opt-out organ donation — dramatically increases donation rates
- Calorie labeling at point of sale — reduces consumption
- Default pension contribution rates — increases retirement savings
- Fly targets in urinals (Amsterdam airport) — reduces spillage 80%
- Fruit at eye level in cafeterias — increases healthy food selection
- Social norm messaging ("90% of your neighbors pay taxes on time") — increases tax compliance
- 21 / 30
Slide 22
Fallacies of Random Sequences
- Probability Intuition
- Two opposing fallacies about random sequences illustrate how our pattern-seeking minds struggle with true randomness:
- The Gambler's Fallacy
- The belief that after a run of one outcome (heads, heads, heads), the opposite is "due." But independent random events have no memory. A fair coin that's landed heads five times has a 50% chance of heads on the sixth flip — exactly as it always did. Casinos know this; gamblers forget it repeatedly.
- The fallacy is strongest for events perceived as controlled by a random mechanism — roulette, dice, lottery. It's weaker when people believe there's a skill component — leading to the opposite fallacy.
- The Hot Hand Fallacy
- The belief that a basketball player who has made several shots in a row is "hot" and more likely to score. Originally identified as a fallacy by Gilovich, Vallone, and Tversky (1985), it was controversially re-examined in 2016 by Miller and Sanjurjo, who found genuine but small hot-hand effects do exist in NBA shooting data — the debate continues.
- The Tension
- Gamblers apply the fallacy to genuinely random events (dice, roulette) but apply the hot-hand belief to genuinely random-with-noise events (sports). The trick is distinguishing truly independent processes from ones with genuine serial correlation. Human intuition is poor at both.
- 22 / 30
Slide 23
The Mere Exposure Effect
- Familiarity & Preference
- Familiarity breeds fondness. Simply being exposed to something repeatedly makes us like it more — even when we cannot consciously recall the previous exposures. Novelty is suspicious; familiarity is comfort.
- Robert Zajonc's Discovery
- Psychologist Robert Zajonc demonstrated in 1968 that people prefer stimuli they've been exposed to more frequently — Chinese characters, faces, nonsense words — even at subliminal exposure levels. The effect works even when subjects don't recognize having seen the stimuli before. Mere exposure, without any additional information, creates preference.
- Advertising Science
- The entire logic of brand advertising — reaching consumers repeatedly without a specific purchase occasion — is built on the mere exposure effect. You don't need to communicate new information; you only need to increase familiarity. "Reminder advertising" keeps brands top-of-mind for future decisions.
- Implications
- People prefer their mirror image to their photograph (mirror = what we see most)
- Popular songs improve with repeated listening before peaking and declining
- Voters prefer familiar-sounding candidate names
- Stocks with easy-to-pronounce ticker symbols outperform immediately after IPO
- Names more frequently printed in newspapers are judged as more historically significant
- Warning
- The effect has limits and can reverse with sufficiently high exposure — explaining why overplayed songs and overfamiliar faces eventually produce aversion. The "sweet spot" of familiarity is culturally and individually variable.
- 23 / 30
Slide 24
The Optimism Bias
- Self-Assessment
- Most people believe they are less likely than average to experience negative events — divorce, cancer, car accidents, job loss — and more likely than average to experience positive ones. This is statistically impossible, but psychologically ubiquitous.
- How Pervasive?
- The optimism bias is one of the most universal cognitive biases — found across cultures, age groups, and levels of education. Even people who know about it display it. Tali Sharot's research identifies it as the most common human cognitive bias: 80% of people exhibit it. It may be the default setting of the human brain.
- The Neuroscience
- Sharot's neuroimaging research showed that when people update their beliefs after negative information, brain regions associated with learning activate less strongly than when updating after positive information. The brain literally processes bad news as less salient than good news — a systematic asymmetry in belief updating.
- Is It Adaptive?
- The optimism bias may serve important functions: optimists are more resilient after failure, more likely to persist, healthier, and live longer. Realistic pessimists are often "depressive realists" — more accurate but less motivated and productive.
- The challenge is calibrating optimism: enough to maintain motivation and resilience, not so much as to make catastrophically bad decisions. Entrepreneurs need it to start companies; they also need to know when to cut losses.
- Unrealistic Optimism at Scale
- Financial bubbles, wars, and mega-projects are often driven by collective optimism bias — shared overconfidence that "this time it's different." The 2008 financial crisis featured systematic optimism about housing prices from buyers, sellers, regulators, and rating agencies simultaneously.
- 24 / 30
Slide 25
The Spotlight Effect
- Social Anxiety
- We overestimate how much others notice and remember our appearance, behavior, and mistakes. We believe we're on stage in a spotlight; in reality, others are too busy managing their own spotlights.
- The Barry Manilow Study
- Thomas Gilovich and colleagues asked participants to enter a room wearing a conspicuous Barry Manilow T-shirt and then estimate how many others in the room had noticed it. Participants consistently overestimated — in one study by more than double. Others were engaged with their own concerns and barely registered the shirt at all.
- Illusion of Transparency
- A related bias: we overestimate how visible our internal states are to others. When nervous, we believe we look more visibly anxious than we do. When lying, we feel more transparent than we are. Our internal experience is salient to us — and we assume it leaks outward more than it does.
- Consequences
- Social anxiety maintained by overestimated audience scrutiny
- Fear of public embarrassment disproportionate to actual scrutiny
- Reluctance to speak up in groups due to exaggerated self-monitoring
- Post-embarrassment rumination disproportionate to actual notice by others
- The Liberating Implication
- Others notice your bad hair day, your stumble, your clumsy comment far less than you fear. Their attention is on themselves. This knowledge doesn't immediately dissolve social anxiety, but it does offer genuine rational reassurance.
- 25 / 30
Slide 26
The Replication Crisis in Bias Research
- Science & Methodology
- In 2015, the Open Science Collaboration attempted to replicate 100 psychology studies — only 39% successfully replicated. This "replication crisis" has forced a reckoning with the reliability of many widely cited cognitive bias findings.
- What Failed to Replicate
- The priming studies in John Bargh's work (elderly words → slower walking)
- Ego depletion (willpower as a limited resource)
- Power posing's hormonal effects (Cuddy) — physical effects replicated; hormonal mechanism didn't
- Some versions of the "stereotype threat" effect
- The "money primes materialism" priming studies
- What Has Replicated Robustly
- Confirmation bias — extremely robust
- Loss aversion — replicated across cultures
- Anchoring — highly reliable, especially in negotiations
- The availability heuristic — well-replicated
- In-group favoritism — Minimal Group Paradigm robust
- Overconfidence bias — consistent across populations
- The replication crisis is not a failure of science — it is science working correctly, detecting and correcting errors. The long-term result is a more rigorous and reliable body of knowledge. Cognitive bias research has emerged more robust, not less.
- 26 / 30
Slide 27
Are Biases Universal?
- Cross-Cultural Psychology
- Most cognitive bias research was conducted on WEIRD subjects — Western, Educated, Industrialized, Rich, Democratic. Growing cross-cultural research shows biases vary significantly across cultures.
- The WEIRD Problem
- Joseph Henrich and colleagues (2010) showed that WEIRD populations are outliers on many psychological dimensions — more individualistic, more analytical, less holistic in thinking. Many studies previously thought to identify universal cognitive tendencies may in fact describe Western patterns. An estimated 96% of psychology study subjects come from only 12% of the world's population.
- Cultural Variation in Biases
- Fundamental attribution error is much weaker in East Asian cultures that emphasize situational thinking
- Optimism bias varies — less pronounced in Japan, where realistic self-assessment is culturally valued
- Social comparison processes differ between individualist and collectivist cultures
- Framing effects show similar direction cross-culturally but different magnitudes
- In-group favoritism patterns shift in cultures with different group-boundary definitions
- 27 / 30
Slide 28
Can We Debias Ourselves?
- Applied Psychology
- The frustrating truth: knowing about a bias does not reliably reduce it. But specific training, system design, and institutional practices can help. The goal is not bias-free thinking but better calibrated thinking.
- What Doesn't Work
- Simply learning that biases exist
- Motivated effort to "think harder"
- Intelligence — smart people show most biases equally
- Experience alone (without calibration feedback)
- What Helps
- Consider-the-opposite strategy — forced generation of alternatives
- Structured decision processes (checklists, red teams)
- Reference class forecasting for planning
- Pre-mortems: "Imagine this failed — why?"
- Actuarial models over expert intuition for high-stakes repetitive decisions
- Calibration training with outcome feedback
- The Intelligence Paradox
- Studies by Keith Stanovich and others show that intelligence (IQ) predicts cognitive bias susceptibility only weakly. High-IQ individuals show most biases at comparable rates — but are often more skilled at constructing post-hoc rationalizations for biased conclusions. Intelligence makes you a better advocate for your position, not necessarily a better reasoner about it.
- Philip Tetlock's Superforecasters
- The best forecasters studied by Tetlock are not those with the highest IQ, but those who treat beliefs as probabilities to be updated, actively seek disconfirming evidence, and view being wrong as information. Epistemic humility combined with systematic process, not raw intelligence, produces calibrated judgment.
- 28 / 30
Slide 29
The Bias Blind Spot
- Meta-Bias
- The meta-bias: people readily recognize cognitive biases in others but are systematically unable to see the same biases in themselves. The very intelligence that helps identify bias in the abstract is used to explain why one's own judgments are exempt.
- "Everyone else is biased. I'm just seeing things clearly."
- Emily Pronin's Research
- Pronin and colleagues demonstrated that people rate their own susceptibility to cognitive biases as significantly lower than average — while simultaneously recognizing that "most people" are biased. The ratings they gave themselves were essentially identical to their rating of an "average American." They saw the bias in others; they were unable to see it in themselves.
- The Motive
- The bias blind spot is driven by motivated reasoning — we are invested in seeing ourselves as rational. When we introspect about our own decision processes, we experience them as logical and considered. We have access to our intentions (which feel rational) but not to the unconscious processes that actually drive our judgments.
- The Recursive Problem
- Even reading this slide has likely increased your sense that you understand biases — and therefore reduced your vigilance about your own susceptibility. Knowledge of bias can paradoxically increase confidence in one's own rationality. Welcome to the bias blind spot about the bias blind spot.
- 29 / 30
Slide 30
Living with Imperfect Minds
- Synthesis
- Cognitive biases are not bugs to be eliminated but features of a mind built for speed over accuracy — in a world that sometimes demands the reverse. The goal is not to become unbiased but to become better calibrated.
- The Core Lessons
- All judgment is influenced by factors we are unaware of
- Intelligence does not confer immunity to bias
- Context, framing, and presentation shape judgment as much as content
- Knowing about biases helps less than systematic processes that correct for them
- The bias blind spot means vigilance must be structural, not just intentional
- The Practical Path
- Design systems and checklists, not just good intentions
- Seek calibration feedback — track your predictions
- Steelman opposing views before dismissing them
- Ask: "What would change my mind?"
- Treat epistemic humility as a skill, not a weakness
- Build institutions that correct for individual bias
- "The first principle is that you must not fool yourself — and you are the easiest person to fool."
- — Richard Feynman
- 30 / 30
