Chapter 1: Heuristics & biases

Misattribution of arousal: Dutton & Aron’s (1974) Capilano Bridge Study

Female researcher approaches men walking across

• Capilano Suspension Bridge

• Solid bridge upriver

The asked them to write a brief story of the picture, and offered her phone number

Results:

• Men are much more likely to call this number than women.

• Men are much more likely to call this number when the researcher is a woman than a man

• Men are much likely to see the picture is a way on the high bridge than a low brick bridge

Limitation: People who choose to cross this bridge might have traits that seek excitement and adventures

What did we say abt social psyc from the capilano bridge study

1 Context matters, often more than we assume

Ex. Where you meet someone could shape whether you’re attracted to them

Importance of subjective perception

2 To predict someone’s behavior, you need to know how they interpret the situation (rather than the objective characteristics of the situation)

• Objective characteristics: Shaky bridge, Super high, above the rocky river

• Subjective perception: Very scary or not, depends on each person

3 Science is a conversation

• Science is NOT a list of facts for you to memorize

• A conversation across time, around the globe

• Goal: Getting closer to the truth

• Science is self-correcting

• The bridge study is pretty weird and flawed, but It was a conversation-starter, not the final word

A tale of two minds

Intuition side: Experiential system, or system 1

Reason side: Cognitive system, or system 2

“A bat and a ball cost $1.10 in total. The bat costs $1.00 more than the ball. How much does the ball cost?”

Intuition first, 80% of harvard students said 10 cents (wrong)

Reason: The bat costs 'x + $1.00'.

The total cost is x + (x + $1.00) = $1.10

2x + $1 = $1.10

The ball costs $0.05 (correct answer)

Key idea: We approach problems in both intuition (experiential system) and reason (cognitive system)

Schemas

Mental maps

Organized set of experiences

Easier to recall the instructions because we have a schema for doing laundry. These vague instructions are mentally organized under one concept

Can shape perceptions of other people (Kelly, 1950: Warm or cold guest prof)

Example of recalling an instruction to use smth. It was hard to recall, until you know that it’s a washing machine, recalling became much easier.

Schemas can alter reality: Self-fulfilling prophecy

Self-fulfilling prophecy

Example of shy roommate: The cycle of Thinking your roomie is shy → don’t talk to roomie much → Roomie thinks you don’t want to talk → Roomie gets nervous and shy → You think they’re shy

Rosenthal’s study: Researcher gave students a fancy IQ test at the start of the school yr.

→ He told the teachers the list of students that “scored high IQ” will “bloom” this year (random asm, so their IQ could be high or not)

→ teachers gave “bloomers” more challenging material, better feedback (this special attention made the children bloom

→ at the end of the yr these students scored higher on an actual IQ test (so they actually bloomed?) due to teachers’ expectations

The example in which a gorilla walked through the room while you counted passes.

Subjective perception

Rating prof study (Kelley, 1950)

Shows schema

Conditions of the guest professor, teach and ask the students to rate this prof

• Expected to be cold: Industrias, critical, practical, determined

• Expected to be warm: Industrious, critical, practical, determined

Expected warm → Much better rating

Conclusion: When situation is ambiguous, we may “fill-in-the-blanks” with info that seems to fit

Think back to the study by Kelly in which a guest professor gave a lecture. This study most clearly illustrates what idea?

The importance of subjective construal: How our expectations and preconceptions can shape our perceptions and evaluations of others.

Availability heuristic (AH)

Judging probability based on how easily examples come to mind

Example 1: Sharks vs. coconut death: Death from sharks seems more common than death from contacts with coconut, even tho there were 4, and 150, respectively

People think that shark attacks are more common because vivid examples of deadly shark attacks easily come to mind

This heuristic helps make judgements based on how easily smth comes to mind. So easy recall → more common (Valuable shortcuts, but could be wrong)

Example 2: Recall 6 times vs. 12 times ppl were assertive: 6 times → assess themselves as more assertive

Example 3: MOP active learning sesh

Biases:

• Retrivability bias (famous men list → assume more men than women)

• Search set effectiveness: Find words beginning with a letter more easily than those with the letter in the middle

• Imaginability bias: Judging based on recalling examples

• Illusory correlation

Asked a group to recall 6 times, and another group to recall 12 times, they behaved assertively, then made themselves assess their assertiveness

AH

Participants who recalled only 6 examples (Wow, I can easily think of times I was assertive!) rated themselves as more assertive than those who had to recall 12 examples (I’m struggling to think of examples..)

Why?: People judged their own assertiveness based on how easily examples came to mind, not the actual number of examples generated.

Anchoring bias (AB)

This bias affects decision-making by heavily influencing estimates/judgments based on the initial piece of info seen, regardless of its relevance.

Biases:

• Insufficient adjustment: People fail to adequately adjust their estimates away from the anchor

• Numerical estimation: Numerical estimate based on first number seen

• Compound probability (conjunction/disjunction fallacies)

• Overconfidence in subjective probabilities

• Calibration issues: Confidence fails to match actual performance, but overrate their accuracy

Iowa Supermarket (Wansink et al., 1998) 10% off sale on soup

Shows anchoring bias

• Limit 12 per person on some days

• No limit per person on some days

Result:

When there’s limit of 12 → people buy 7 cans

Twice as many as when there’s no limit

German judges (English et al., 2006)

Illustrates the anchoring bias, showing how random information, like dice rolls, can skew decision-making processes, even among experienced professionals.

15 years experience

Read a shoplifting case

Roll a pair of (weighted) dice

• Always roll 3 or 9

• Sentence to more or fewer months than dice roll

Judges decide on sentence

• Roll a 3: 5 months

• Roll a 9: 8 months

People got 3 more months of probation just because of the dice. The anchoring effect shows that even irrelevant or random numbers can unconsciously influence decisions, including those made by trained professionals like judges.

Representativeness heuristic (RH)

Judging probability based on how much an example resembles a known category or stereotype

Key biases:

• Insensitivity to base rates

• Insensitivity to sample size: Ppl think small sample’s representativeness = large sample

• Misconceptions of chance (H/T)

• Insensitivity to predictability

• Illusion of validity

• Misunderstanding regression toward the mean

Insensitivity to base rates

RH

People ignore statistical quantities of ppl in a career (e.g., more farmers than librarians) and judge likelihood purely by stereotype match. This bias leads individuals to overlook the actual probabilities of events, focusing instead on how well a specific example fits into a given stereotype.

Engineer vs. lawyer study

Insensitivity of base rates (RH)

Participants were told about a group of 100 individuals, either:

• Group A: 70 engineers and 30 lawyers, or

• Group B: 30 engineers and 70 lawyers.

Then they were presented with short descriptions of individuals from these groups.

• Description of highly stereotypical of either engineers (introvert, like math) or lawyers (extrovert, assertive): People tend to ignore the disproportion of the career groups if the description sounds like a typical member of one group

• Description of neutral or uninformative (John is a 30 yr old man, married with no children. High motivation and likable) – Even though base rates should dominate the judgment in the absence of new info, people often default to 50/50 or guessed randomly.

Insensitivity to sample size

RH

People think small samples are just as representative as large ones. Small samples are more likely to produce extreme outcomes due to greater variability

Babies in hospital Study: Participants judged the likelihood of sample results (e.g., babies born in small vs. large hospital) as equal, ignoring that smaller samples have more variability

Misconceptions of chance

RH

People expect small sequences (like coin tosses) to look random (e.g., alternating heads/tails), leading to the gambler’s fallacy.

Coin tossing Study: Sequences like H-T-H-T-T-H seen as more likely than H-H-H-T-T-T

Insensitivity to predictability

RH

People make CONFIDENT predictions based on descriptions even if the evidence is unreliable.

Example study about teacher’s success: Participants rated the likelihood of a teacher's future effectiveness based on a single lesson, ignoring broader context and variability in performance.

This phenomenon occurs when individuals underestimate the actual predictive validity of particular indicators or information, leading them to draw overly confident conclusions without sufficient evidence.

Illusion of validity

RH

CONFIDENT in predicting abt smth based on how well information "fits" the outcome, even when the info is weak or irrelevant.

Psychologists trusting personality descriptions to predict job success, despite knowing interviews are unreliable

Example study: Participants believed they could predict job performance based on personality traits, overlooking the lack of empirical support for such assessments.

Misunderstanding regression toward the mean

RH

People create causal stories for natural performance fluctuations.

Pilot trainees study: The instructor noted that praise for an exceptionally smooth landing is typically followed by a poorer landing on the next try, while harsh criticism after a rough landing is usually followed by an improvement on the next try. The instructors thought it was cuz of them, but trainees are actually in the regressing toward the mean

Retrievability bias

AH

More famous or recent examples seem more frequent.

Famous ppl list study: Lists with famous men than women led students in class to judge there were more men, even when there weren’t.

Search set effectiveness

AH

Influenced by the ease of recalling examples based on their position in words, leading to biased perceptions of frequency.

Words with R Study: People think there’s more words starting with “R” compared those that ends with “R” (even though the latter is more common) cuz we find the word that start with R easier.

The judged frequency/variability of abstract words was much higher than concrete words, equated in objective frequency.

Salience (being important to/connected with what’s happening): Seeing a house on fire urself makes u think it’s more common than reading news abt a house being on fire

Imaginability bias

AH

The easier it is to mentally generate examples, the more likely we judge them to be.

Study: Small committees (2 people) judged as more numerous than large ones (8 people) due to ease of imagining

Example study: Participants judged the occurrence of natural disasters as more frequent because they could easily recall instances from recent news.

Illusory correlation

AH

People overestimate the co-occurrence of two related traits.

Sus Study: Subjects believed "suspiciousness" was linked to drawings with "peculiar eyes," reflecting clinical stereotypes. There’s no correlation whatsoever, but ig sus is readily associated w eyes than other parts of the body.

Race vs. criminal behavior study: No real pattern between nationality and criminal behavior, but people mistakenly perceive a connection.

Insufficient adjustment

AB

Initial information (an anchor) influences subsequent judgments or estimates, even when that information is irrelevant

Study: Wheel-of-fortune spun to 10 vs. 65 changed estimates of % of African countries in the UN

Numerical estimation

AB

Provide estimates of quantities based on an initial numerical reference point (first number read), often leading to skewed results.

Study: Students estimated the product of descending numbers (8×7×6...) as higher than ascending ones (1×2×3...), despite the values being equal, due to early-step anchoring.

Biases in compound probability (conjunction/disjunction fallacies)

AB

Overestimate conjunctive events (relating to connections between things), and underestimate disjunctive events (lack clear connections). Fail to recognize the actual probabilities involved in each scenario.

Marble Study: Betting on 7 red marbles in a row (p = .48) seemed more likely than a single draw (p = .50), which is mathematically incorrect.

Overconfidence in subjective probabilities

AB

People give confidence intervals that are too narrow.

Overestimate the accuracy of their knowledge or predictions.

Calibration issues due to anchor choice. Best guesses serve as anchors; insufficient upward/downward adjustment leads to overconfident estimates.

Example study: Subjects expected 90% of values to fall within their ranges, but only ~70% did.

Imagine that your classmate tells you that they did this week’s reading before watching the videos, and they found the reading confusing. You remind them of Dr. Dunn’s recommendation to do the reading AFTER watching the videos. What example or study should you remind them about to support this recommendation?

a. The example of believing that the new roommate Suzie is shy

b. The study in which teachers were told that some students would bloom

c. The example of recalling a set of instructions with or without seeing a picture of a washing machine

d. The study in which men were approached by a female experimenter on a stable or shaky bridge

c. The example of recalling a set of instructions with or without seeing a picture of a washing machine

Cuz a. and b. are self-fulfilling prophecy, and d. is misattribution of arousal, but highlights why social psyc matters — 1)context matters, 2)importance of subjective perception, 3)science is a conversation

Ben has 5 participants per week, Eva has 20 per week. One of them, found IQ avg of 140, and another one found IQ avg of 123. Guess.

Ben’s participants had an avg IQ of 140.

Smaller sample size leads to more extreme avg., larger sample size leads the avg. closer to the mean.

Think back to the study by Kelly in which a guest professor gave a lecture. This study most clearly illustrates which of the following ideas?

a. Misconceptions of chance

b. Misconceptions of regression

c. Science is a conversation

d. The power of the situation

e. The importance of subjective construal

e. The importance of subjective construal

Lin’s TA tells all of his students that they should spend about 30 minutes on each problem set. Kia’s TA tells all of her students that they should spend about 2 hours on each problem set. As a result, Lin typically spends 45 minutes on each problem set, while Kia spends 90 minutes.

What study you can relate this to?

a. The study in which people recalled 6 or 12 times they behaved assertively.

b. The study in which German judges read about a shoplifting case.

c. The example in which a gorilla walked through the room while you counted passes.

d. The study in which men were approached by a female experimenter on a stable or shaky bridge.

e. The study in which people judged whether personality descriptions belonged to engineers or lawyers.

b. The study in which German judges read about a shoplifting case (anchoring bias)

a) is availability heuristic, d) is misattribution of arousal, c) is subjective perception, and e) is representative heuristic