Class Presentation

Report
The Psychology of Judgment and
Decision Making
CH1: Selective Perception
What is selective perception
• People selectively perceived what they expect
and hope to see.
• Cognitive factors:
prior belief
expectations
• Motivational factors:
hope
desire
emotional attachments
Calling a spade a spade
Present people with five playing cards on a tachistoscope.
One of the cards is actually a black three of hearts. (Bruner
& Postman, 1949)
Most reactions to incongruity:
Dominance: three of hearts or spades
Compromise: mixed
Disruption: neither
Recognition: something wrong
Conclusion: expectation can
strongly influence perception.
Potent Expectation
• Alcohol research:
• Believed Vodka+tonic -> higher heart rate
• Believed Tonic -> lower heart rate (Wilson &
Abrams, 1977)
• Conclusion: heart rate were not significantly
affected by whether subject had been given alcohol
to drink; they were affected by whether subjects
believed they had been given alcohol to drink.
High sensation seeker
Low sensation seeker
Alcohol (belief)
Aggressive
Cautious
Non-alcohol (belief)
Normal
Normal
Football team case
• Dartmouth and Princeton football
team got conflict in a game.
• After watching the video, Dartmouth
students see more infractions on
Princeton team, while Princeton
student see more infractions on
Dartmouth team.
 The “game” actually was many
different games.
 The “thing” different people see is
different. (Hastorf & Cantril)
The hostile media effect
• Most people believe the media coverage of candidate who
they support are biased.
• Similar biases in perception might arise in the context of
mediation, arbitration , or other situations in which two
sides are heavily committed to prior positions.
Perceptions are
selective.
Decision makers should
be careful of biases in
perception.
CH2: Cognitive Dissonance
What is cognitive dissonance
• After selective perception, Leon Festinger (1957)
proposed the theory of “cognitive dissonance.”
• People are usually motivated to reduce or avoid
psychological inconsistencies.
Boredom can be fun
• in the experiment, subjects were asked to do boring tasks and
to tell others the task was interesting. Some got 1$ and some
got 20$.
• After the experiment, more subjects who got 20$ reported
the task was boring than who got 1$.
 1$ subjects suffered from cognitive dissonance, so they
changed their options to avoid it.
 20$ subjects got the payoff for telling a lie, no cognitive
dissonance.
Self-perception theory
• Another way to account for what Festinger found
• Dissonance findings have to do with how people
infer their beliefs from watching themselves behave.
• Two main premises:
• 1. people discover themselves by watching their
behaviors.
• 2. internal cues are weak, ambiguous, or
uninterpretable.
Predicisional dissonance
• People tends to behave more “liberated” (try to show they
are not sexist) after noticing that they have sex-role
stereotypes. (Sherman & Gorkin, 1980)
• Heavy users of electricity cut their consumption significantly
when they were informed of their heavy use and reminded of
an earlier conservation endorsement they had made. (Kantola
et al., 1984)
• Customer “adaptation levels” and the need to avoid
dissonance. (Doob, 1969)
Postdecisional dissonance
• After placing a bet on
horse racing, the bettors
tend to believe horses
they support are more
likely to win.
• Voting a candidate
increase your confidence
that the candidate will
win the election.
conclusion
• Cognitive dissonance is applicable to many
situations. (political campaigns, retailing)
• Both cognitive dissonance theory and selfperception theory can explain cognitive
dissonance phenomena.
• Change in attitude->change in behavior
• But, change in behavior->change in attitude
Chapter 3: Memory and
Hindsight Biases
Memory
• IS NOT
– What exactly happened in the past, or authentic
copies of past experiences
• IS
– Something constructed at the time we are
recalling
– Something that people fill in missing details with
logical inferences and associated memories
Interesting Example
• When people viewed accident film clips, their
estimates of car speed varied based on how the
question was worded (How fast were the cars going
when they “hit” verses “smashed” each other?)
• People who were asked about the cars “smashing”
each other also remembered seeing broken glass
(but there was no broken glass)
“Smash” =>
Hindsight Bias
• “Is the inclination to see events that have
occurred as being more predictable than they
were before they took place.”
• People are subject to hindsight bias (I knew it
all along)
• How to avoid hindsight bias
– Besides the real outcome,
try to find all the other outcomes
as well as the factors cause this outcome
Lesson learned
Don’t fully believe everything we remember, and
using memory cautiously when we are making
decision and judgment
Record important details for the future reference
Chapter 4: Context Dependence
Context Dependence Effects
Four types of context dependence effects:
1. The Contrast Effect - Comparison of objects to
other similar objects influences our perception
2. The Primacy Effect - First impressions affect our
judgments more than later impressions
3. The Recency Effect - Our judgments can be more
affected by things we heard or saw recently
4. The Halo Effect - Favorable impressions of one trait
result in increased impressions of other traits
Example
• Contrast Effect :
– Put your hand into water with different
temperature
– Height of sports announcer
Example
• Primacy effect
– When people were given a list of characteristics of
someone, the items early in the list affected their
judgment of the person more than the items later
in the list
“He is
Warmhearted Kind
,
selfish…”
This guy most likely will be marked as “Warmhearted”
, introverted,
Example
• Recency Effect :
– When subjects were presented at court, their
judgments reflected a recency effect when there
was a delay between hearing the two sides of the
case
Choose a better time slot to present idea
TimeSlot1
TimeSlot2
TimeSlot1
Delay
Delay
TimeSlot2
Response
Primacy effect
Response
Recency effect
Example
• Halo effect:
– Flight commanders showed a correlation between
ratings of subordinates’ intelligence and physique
Lesson learned
Our judgments and decisions are not context-free
Take advantage of context effect (such as how to
select good time to present idea)
Plasticity
Chapter 5 and Chapter 6
Guideline
• Plasticity
• Pseudo Opinion
• Inconsistency
Plasticity
Context
Wording
Order
Response
wording
• A. A 100% chance of losing $50
• B. A 25% chance of losing $200, 75% chance of losing
nothing.
$50 to buy an
insurance ?
Pseudo Opinion
opinion on a topic about which
people actually have no real
opinions.
Sometime it matters!
Inconsistency
• Attitude-attitude inconsistency
– Discrepancy between two related attitudes
• attitude behaviour inconsistency
– Discrepancy between attitude and corresponding
behaviour
Conclusion
• Attitudes, opinions and choices are very
plastic.
• Measuring an attitude, opinion or preference
is not simple.
• It is worth paying attention to the structure
and context of questions.
The effects of question wording
and framing
Guideline
• Examples of different way to ask questions.
• Framing
• Conclusion
Middle category
Slower
Slower
Faster
Keep
Faster
Open vs closed question
Structured and closed response
alternatives are never perfectly
neutral.
Social desirability
• People tend to answer in the way that is more
socially desirable.
'Allow' vs 'Forbid'
• Do you think US should allow anti-democratic
speech? 62% said No
• Do you think US should forbid anti-democratic
speech? 46% said Yes
Framing
• A decision frame is the
decision maker's
conception of the acts,
outcomes and
contingencies
associated with a
particular choice.
Conclusion
• Question wording and framing often make a
substantial difference in answers.
• Thus it pays to be aware of the effects.
• Safest way is to elicit answers in different ways,
and then compare the results.
Question
• Which is true about framing?
• A. A frame is the decision maker's
conception of styles and alternatives
associated with a choice
• B. Both choices and outcomes can be
framed
• C. In most cases, frames are dominated
by norms and habits of the decision
maker.
• D. All of above
Section 3: Models of Decision Making
Chapter 7
Expected Utility Theory
St. Petersburg Paradox
An unbiased coin appears first tail at
Odds
Money earned
1st toss
1/2
$2
2nd toss
1/4
$4
3rd toss
1/8
$8
…
Expected Value (
)=
: How much would you pay to play the game?
Nicolas Bernoulli
Marginal Utility
U ( game)   pi  u( xi )
u' ( xi )  ?
• Classic Utility Theory
• Expected Utility Theory
Expected Utility Theory
• Principles of Rational Decision Making
– Ordering of alternatives: Prefer one alternative or indifference
– Dominance: Comparison of different attributes
• Weak vs. Strong
–
–
–
–
Cancellation: Common factors contribute nothing
Transitivity : If (A > B) & (B > C), then (A > C)
Continuity: Given L ≤ M ≤ H, ∃ p: M ≤ pL + (1-p)H
Invariance: Toss one coin twice is the same as toss two coins once
• Extensions (Expected Utility Theory is a family of theories)
– Subjective expected utility theory: For unprecedented evens
– Stochastic models of choice: For random evens
Chapter 8
Paradoxes in rationality
The Allais Paradox
A1 = Certainty of $1 Million
A2 =
B1 =
10
89
1
$1M
$1M
$1M
1/100 Chance of $0
89/100 Chance of $1 Million
10/100 Chance of $5 Million
10
89
1
$5M
$1M
$0M
89/100 Chance of $0
11/100 Chance of $1 Million
10
89
1
$1M
$0M
$1M
10
89
1
$5M
$0M
$0M
90/100 Chance of $0
B2 =
10/100 Chance of $5 Million
Ellsberg’s Paradox
• Let’s draw balls again!
30 Balls
60 Balls
Betting Alternatives
Red
Black
Blue
A1: a red ball
$100
$0
$0
$0
$100
$0
A2: a black ball
30 Balls
60 Balls
Betting Alternatives
Red
Black
Blue
B1: a red ball
$100
$0
$100
$0
$100
$100
B2: a black or blue ball
Intransitivity
• If (A > B) & (B > C), then (A > C)
• What if:
Case 1:
*Decision Rule*
If the difference in IQ between
any two applicants is greater
than 10 points, choose the more
intelligent one. If the difference
between applicants is equal to
or less than 10 points, choose
the one with more experience.
Case 2:
APPLICANTS
Gamble
A
B
C
D
E
A
B
C
DIMENSIONS
Intelligence (IQ) Experience, Years
120
1
110
2
100
3
Probability of a Win
7/24
8/24
9/24
10/24
11/24
Payoff, $
5.00
4.75
4.50
4.25
4.00
EV, $
1.46
1.58
1.69
1.77
1.83
Preference Reversals
• Lottery A:
– 9/10 chance of winning $100 and 1/10 chance of losing $10
• Lottery B:
– 1/10 chance of winning $1000 and 9/10 chance of losing $100
Findings from several studies:
When people are asked to choose between the two bets, they pay
particular attention to the probability of winning, but when they are
asked to set a price of how valuable the bet is, they look as how large
the potential payoffs are.
You might ask…
• Are violations of expected
utility theory truly irrational?
• (Almost) certainly…no.
– Why? Because we have no
information about the cost of
people’s errors compared with
the cost of following the
principles.
Chapter 9.
Descriptive Models of Decision Making
Are we economically RATIONAL?
Location
Price
Space
Safety
People “satisfice” rather than “optimize”
when they make decisions.
Prospect theory
Daniel Kahneman
Utility
Value
Amos Tversky
Example
Alt. A
A 50% chance of gaining $1,000
Alt. B
A sure gain of $500
$1,000
84 percent
(70 respondents)
70 percent
(68 respondents)
$2,000
Alt. C
A 50% chance of losing $1,000
Alt. D
A sure loss of $500
Extending….
Refers to the tendency to give greater weight to certain
outcomes than to uncertain outcomes.
Certainty
Effect
Prospect
Theory
Pseudocertainty
Effect
Regret
Theory
Regret
People compare the quality of their decision
in terms of ‘
’ and ‘
’.
Refers to make risk-averse choices if the expected outcome
is positive, but make risk-seeking choices to avoid negative
outcomes.
Example of Pseudocertainty Effect
40 %
A vaccine that protected half the recipients
Probabilistic
from a disease that was expected to afflict
protection
20 percent of the population
57 %
Pseudocertainty
A vaccine that complete protection against
one of two mutually exclusive and equally
probable strains of the disease, each of
which was expected to afflict 10 percent of
the population.
Lessons learned
• There are biases and interpretive strategies
that shape people’s choices in the face of
uncertain outcomes.
• Prospect theory helps to interpret
experimental results that demonstrate
individuals often make divergent choices in
situations that are basically identical but
framed in different ways.
Section 4:
Heuristics and Biases
Chapter 10,11,12,13
Heuristics and Biases
• Complicated decisions -> rules of thumb
• Some biases & inconsistencies -> predictable
Representativeness Heuristic
Number of bank tellers > Number of feminist bank tellers
Nonregression Prediction
Student Percentile
Achievement Test
GPA
Top 10%
> 750
> 3.7
Top 20%
> 700
> 3.5
Top 30%
> 650
> 3.2
Top 40%
> 600
> 2.9
Top 50%
> 500
> 2.5
GPA
4
3.5
3
GPA
Log. (GPA )
2.5
2
500 525 550 575 600 625 650 675 700 725 750 775
Conclusion
•
•
•
•
Don’t be mislead by highly detailed scenarios
Pay attention to base rates
Chance is not self-correcting
Watch for regression toward the mean
Availability Heuristic
Availability can lead to bias
Power of Vividness
The Power of Vivid Testimonials
Condition
Recommended Course
Non-Recommended
Courses
Face-to-face
4.73
0.50
No evaluation (control)
3.33
1.39
Base rate
4.11
0.94
The Legal Significance of Guacamole
• On his way out the door, Sanders [the defendant] staggered against a table,
knocking a bowl on the floor.
• On his way out the door, Sanders [the defendant] staggered against a table,
knocking a bowl of guacamole dip to the floor and splattering guacamole on the
white shag carpet.
Confusion of the Inverse
Car is in door:
You choose:
Host opens:
You switch to:
Result:
1
1
2 or 3
2 or 3
Lose
1
2
3
1
Win
1
3
2
1
Win
2
1
3
2
Win
2
2
1 or 3
1 or 3
Lose
2
3
1
2
Win
3
1
2
3
Win
3
2
1
3
Win
3
3
1 or 2
1 or 2
Lose
Switching doors is a better strategy!
Right Monty?
Compound Events
People prefer compound bets over simple bets
It’ll Never Happen to Me
Students rate themselves against their peers:
42% more likely for a better salary
58% less likely for drinking problem
Pro vs. Anti Subjects
“Several serious human and technological breakdowns have taken
place without leading to nuclear war.
Do such breakdowns give you greater confidence that inadvertent
nuclear war will occur, less confidence, or neither?”
Recommendations
• Watch for availability, vividness, inverses
• Beware of wishful thinking
• Break compound events into simple events
Anchoring
Listing Price, $
Reasonable Purchase Price, $
120,000
111,000
130,000
123,000
140,000
125,000
150,000
127,000
Chapter 14,15,16
Coincidence
• Cases
– George D. Bryson : You’ve got mail
– Intel and infringement
– German Mother: WW II
Is it an Act of
God????
Perhaps or Perhaps Not…..
What can be empirically investigated, are the
answers to two questions
1. Do people tend to see meaningful patterns in
random arrangements of stimuli?
2. Can people behave randomly?
Luck and Superstition
• 40% of Americans believer that some numbers
are especially lucky for some people
• Claims to be lucky in future - Superstition
• Experiment by Harold Hake and Ray Hyman
(1953)
• People perceive an ambiguous series of events
as being more structured that it already is
Randomness
•William Wagenaar’s
study on people seeing
patterns in randomness
•People have difficulty
generating random
sequences
•Learning to act
randomly
Lesson Learned
• Practical implications of research on
perceived randomness - Tendency to over
interpret chance events
• Easy to see patterns in random outcomes
• Decision makers should resist viewing the
short runs of same outcome as meaningful
Co-variation Assessment
• Assessment of whether two variables are
correlated
• Does God answer prayers?
Event
Happened
Pray
Didn’t Pray
Event Did
not Happen
Illusory and Invisible Correlation
• Illusory - The mistaken impression that two
unrelated variables are correlated
• Draw-A-Person Example
• Invisible - Failing to see correlation between
two variables, due to absence of expectation
Causalation
• Misinterpreting correlation as causation
“Correlation doesn’t imply causation, but it does waggle its eyebrows
suggestively and gesture furtively while mouthing look over there”
• Causalation – Belief that causation implies
correlation
Lesson Learned
• Correlation assessments are simply
generalizations based on previously
established relationships
• Decision maker must weigh new and
conflicting factors in information to arrive at a
fresh assessment.
• Focus more, What did not take place is equally
significant as what did
• Distinguish between correlation and causation
Attribution Theory
A psychological theory about how people make
"causal attributions" or explanations for the
causes of actions and outcomes.
Expressing behavior
Person
Entity/Stimulus
Circumstance
Consensus, Distinctiveness, Consistency
Predictions from Attribution Theory
Predicted
Attribution
PATTERN OF INFORMATION
Consensus
Distinctiveness
Consistency
Person
Low
Low
High
Stimulus (Entity)
High
High
High
Circumstance
Low
High
Low
Adapted from an article by Bruce Orvis, John Cunningham, and Harold Kelly (1975)
Salience
Information that is salient, available, or vivid
tends to have more impact than information
which is not.
Attribution Biases
• People disregarding consensus information –
following dispositional factors
• Self Serving Bias – People are more likely to accept
responsibility for successes than for failures
• Egocentric Biases – People accept more
responsibility for joint outcomes than other
contributors attribute to them
• Positivity Effect – Tendency to attribute positive
behaviors to dispositional factors and negative
behavior to situational factors
Fundamental Attribution Error
Over readiness to explain behavior in terms of
dispositional factors
↓their actions to
Actors attribute
Most salient
feature during
a social setting is
situational
requirements
actor’s behavior and not the situation
Lessons Learned
• Avoid Attributional Bias
• Ask how you would have behaved if in the
same circumstance
• Look for hidden causes
Section V – Social Side of
Judgement and Decision Making
Ch 17: Social Influences
Ch 18: Group Judgements and
Decisions
Chapter 17: Social Influences
• Social factors play pivotal role in judgment and
decision making
• Phillip Tedlock (1985b)
Social Facilitation
• Simple, well-learned tasks
are enhanced by onlookers
• Complex, unmastered
tasks are impaired
15%
10%
5%
– Robert Zajonc (1965)
0%
Above Average
• Example-College Pool Hall
– Michaels, Blombel,
Brocato, Linkous & Rowe
(1982)
-5%
-10%
-15%
Below Average
Social Loafing
• People don’t work as
hard in groups as they
do alone
– Walther Moede (1927)
• Example-Noise from
clapping and shouting
– Latane, Williams and
Harkins (1979)
9
8
7
6
5
4
Shouting
3
Clapping
2
1
0
1
2
4
Group Size
6
Bystander Intervention
• Bystanders witnessing a
problem are less likely
to act when other
bystanders are around.
• Example- Reporting
Smoke pouring from
vents
– Latane and Darley
(1969, 1970)
Alone
Reported
Not reported
3 Ind. Bystanders
Reported
Not reported
1 with 2 Bystanders
Reported
Not reported
Social Comparison Theory
People…
• …will evaluate their
opinions and abilities
• …compare themselves to
others when objective
information is unavailable
• …prefer comparisons with
others who are close to
them in opinions and
abilities
• Festinger (1954)
• Example-Lost Wallet
80%
70%
60%
50%
Positive Tone
40%
Neutral Tone
Negative Tone
30%
20%
10%
• Hornstein, Fisch, Holmes (1968)
0%
Dissimliar
Similar
Social Conformity
• Many people will conform
to objectively wrong
belief if enough people
around agree
Percent of conformity
40%
35%
30%
25%
• Example- Size of Lines
• Asch (1951)
20%
15%
• Lone dissenters can cut
conformity to ¼ of previous
levels
10%
5%
0%
1
2
3
4
8
10-15
Minority Influence
• A consistent minority
group members can
influence the majority
• Example- Blue/Green
Study
Minority Influenced
Green
Blue
Uninfluenced
• Moscovici, Lage, &
Naffrechoux (1969)
Green
Blue
Group Think
• “a deterioration of mental efficiency, reality testing and moral
judgment that results from in-group pressures”
•
Janis (1982)
8 Symptoms of GroupThink:
• Illusion of invulnerability
• Rationalize or discount
warnings
• Unquestioned group
morality
• Stereotypes of adversaries
• Pressure on dissenters
• Illusion of unanimity
• Self-censor deviations
from group consensus
• “Mindguards” protect
from new information
Chapter 18
Group Judgments and Decisions
Chapter Outline
•
•
•
•
•
Group Errors and Biases
Group Polarization
Horse Sense
Are Several Heads Better Than One?
The Benefits of Dictatorship
Group Errors and Biases
• Stupid Human Tricks
• Group Attribution Error
– Dispositional Attributions
• Group Serving Bias
– Situational Attributions
• Outgroup Homogeneity Bias
Group Polarization
• Choice Shift
• Wisdom of Crowds - Riots
Horse Sense
• A man bought a horse for $60 and sold it for
$70. Then he bought it back for $80 and sold it
for $90. How much did the man make?
• Individuals – 45% correct
• Inactive leader – 72%
• Permissive leader – 84% (Voice)
• 63 of 67 groups contained 1 individual with
the correct answer
Are Several Heads Better Than One?
• Quantities and Magnitude
– Groups Win (23-32% more accurate)
• Brain teasers and individuals
– Best individual outperforms group
• General Knowledge Questions
– Best individual outperforms group
• Wisdom of Crowds – Uncertain Situations
• Hill Brainstorming as Individuals
The Benefits of Dictatorship
• Consensus, Dialectic, Dictator, and Delphi
Outperform Collective Decisions
• The Dictator technique reduced absolute
percent error by three times more than that of
any other technique
• Wisdom of Crowds – This is an uncertain
decision
Chapter 19
OVERCONFIDENCE
60
50
%
40
30
20
10
0
less
more
Information
Accuracy
Confidence
Stuart Oskamp (1965)
Overconfidence
Underconfidence
Confidence Level
0%
20%
40%
Accuracy
60%
80%
100%
Lichtenstein & Fischhoff (1977)
Money Incentive
Can overconfidence
be overcome?
Task Importance
Expertise
Life
Stop &
Think
Chapter 20
Self-Fulfilling Prophecies
“The self-fulfilling prophecy is, in the
beginning, a false definition of the
situation evoking a new behavior
which makes the originally false
conception come true.”
Robert Merton, 1948
#1 - Card Experiment (Wason & Johnson-Laird, 1972):
Vowel on one side  Even Number on Opposing Side
E
K
4
7
#2 - Guess Again (Wason 1960):
Prove 2, 4, 6 rather than disconfirm 6, 4, 2
#3 – Self-Perpetuating Social Beliefs (Snyder & Cantor, 1979):
“…spoke freely with strangers while
jogging…”
“…shy and timid at the
supermarket…”
Will Jane be more suitable for a job in real estate sales or a job at the library?
#4 – The Pygmalion Effect (Rosenthal & Jacobson, 1968):
The results of this study have contributed to 400 or more similar experiments, 100 of which
analyzed the Pygmalion Effect in a learning environment.
#5 - In The Minds of Men – (Snyder, Tanke & Berscheid, 1977):
Study
Men and women instructed
to call participants of the
opposite sex.
Pictures were taken of each
participant, however, these
pictures were not actually
given to their phone partner.
Experiment
#6 - Self-Fulfilling Racial Stereotypes
(Word, Zanna & Cooper, 1974):
***People reveal attitudes
through nonverbal behavior.
Purpose:
Is there a difference in the way
that white interviewers treat
white and black interviewees?
Measures
Interview Length
Errors Committed
Chair Distance
Results
• White interviewers…
– …spent 35% more time with white interviewees
– …committed 50% more speech errors
– …and…
Follow-up Study
• White interviewers were then tasked with
interviewing 50% of interviewees as if they
were black, and 50% as if they were white.
• Interviewees treated as if they were black
felt…
– poorly about their interviews
– had an increased rate of speech errors
– thought that the interviewer was less friendly
• A legitimate self-fulfilling prophecy!
Conclusions
• People have the tendency to seek confirming
evidence.
• Confirmation biases are difficult to eliminate.
– (Subjects received directions to confirm,
disconfirm, or test a phenomenon. 70% still
attempted to confirm it.)
• We can change this behavior by introducing
new motivating factors.
• Frame questions in a way that encourages
disconfirming answers.
Behavioral Traps
Chapter 21
Time Delay Traps
Ignorance Traps
Hard to avoid, because you don’t know the consequences at the beginning
Investment Traps
What if you got the tickets for free?
Deterioration Traps
• Initial benefits turn into costs
– Drug addiction
– Pesticide
Collective Traps
• Pursuit of individual self-interest results in
worse outcomes for all involved
• Two types
– Tragedy of the commons
– Prisoner’s Dilemma
Tragedy of the Commons
Prisoner’s Dilemma
Escaping the Traps
• Take a step back and try to take an objective
look at the situation
• Ignore the effects of sunk costs

similar documents