the psychology of judgment and decision making – scott plous doug derrickaaron elkins shaokun...
TRANSCRIPT
The Psychology of Judgment and Decision Making – Scott Plous
Doug Derrick Aaron ElkinsShaokun Fan Noyan IlkKunpeng Zhang David Zimbra
2
Agenda
Book OverviewSection Reviews
Perception, Memory and Context – KunpengHow Questions Affect Answers – David Models of Decision Making – Aaron Heuristics and Biases – Doug The Social Side of JDM – Noyan Common Traps – Shaokun
Conclusion
3
Book Overview
As scientists, it is imperative that we continually monitor our decision making processes as it is intrinsically subject to bias and errors. We must also be aware how experimental subjects make decisions in order to set up good experiments and conduct valid research.
4
SECTION 1
PERCEPTION,
MEMORY,
AND
CONTEXT
5
Section 1 – Perception, Memory, and Context
All judgments and decisions rest on the way we see, interpret and remember the world
6
Section 1 – Perception, Memory, and Context
Selective PerceptionCognitive DissonanceMemory and Hindsight BiasesContext Dependence
7
Selective Perception
Perceptions are selective and depend critically on cognitive and motivational factorsCognitive – what people
expect to see (prior beliefs and expectations)
Motivational – what people want to see (hopes, desires, and emotional attachments)
8
Selective Perception
Four types of most reactions to incongruity
Potent expectationsThe ‘same’ game is not the
‘same’Hostile media effect
9
Selective Perception
Nearly impossible to avoid biases in perception
Ask yourself questions before making an important decision
10
Cognitive Dissonance
People are usually motivated to reduce or avoid psychological inconsistencies People experience cognitive
dissonance when they simultaneously hold two thoughts that are psychologically inconsistent
“negative driven state” – motivational theory
“self-perception theory” – inferring beliefs from watching behaviors
11
Cognitive Dissonance
Predecisional dissonance – influence the decisions people make
Postdecisonal dissonance – its reduction has an effect on later behavior
Changes in attitude Changes in behavior
Well known
Cognitive dissonance
12
Cognitive Dissonance
Promote energy conservationApplication on electionsThe pressure to feel consistent will often
lead people bring their beliefs in line with behavior
13
Memory and Hindsight Biases
Memory is reconstructive and highly dependent upon contextual factorsShattered memoriesSweet remembrancesHindsight biases –
“I-knew-it-all-along”
14
Memory and Hindsight Biases
Memories are not like copies of our past experiences on deposit in memory bank
People construct and memorize a general scenario
Hindsight biases diminish (consider why and how an alternative outcome might have occurred)
15
Memory and Hindsight Biases
People have difficult telling how they are effected by information about an outcome
The value of keeping accurate records
Decision maker is
susceptible to biases
in memories
16
Context Dependence
Decision makers do not perceive and remember material in isolation;
they interpret new information
in light of past experience
and the context in which
the material occurs
17
Context Dependence
The contrast effect – Ebbinghaus illusionThe primacy effect – first impression delay
delay
The recency effect – final presentationHalo effects – compound and magnitudeThe effects of context are not
limitless
response
response
Primacy effect
recency effect
18
Context Dependence
Context effects are so commonApplication of context dependence Time selection
19
SECTION 2
HOW QUESTIONS
AFFECT
ANSWERS
20
How Questions Affect Answers
Format of the problem influences responsesHow much do decisions depend on the way
a problem is worded?What are specific ways in which the wording
of a question influences the answer?
21
How Questions Affect Answers
PlasticityOrder EffectsPseudo Opinions Inconsistency
Question Wording and FramingWording of the
QuestionResponse Alternatives
Framing
22
Plasticity
Plasticity defined The degree to which responses can be affected
by the wording or presentation of the question or response alternatives
23
Plasticity
Order Effects Responses influenced by
Ordering of response alternatives Asking questions to influence later responses
24
Plasticity
Pseudo Opinions Respondents have no opinion or knowledge
about the topic or issue in question Exhibit ultimate plasticity Offer pseudo opinion in lieu of personal opinion
Influence on political issues
25
Plasticity
Filtering Pseudo Opinions Filters identify respondents
who have no opinion or knowledge about the topic
Asking preliminary questions about knowledge
Offering ‘no opinion’ or ‘don’t know’ response alternatives
26
Plasticity
InconsistencyDiscrepancy between two related attitudes
or an attitude and corresponding behaviorRelated attitudesAttitude and corresponding behavior
27
Question Wording and Framing
Question Wording and Framing Changes of only a word or two can influence
how respondents answer Question Response alternatives
?
28
Question Wording and Framing
Wording of the QuestionMissing middle category
Force the choice of an extremeMiddle category collects 10 – 40%
Open-ended vs. close-ended questionsDifferent responses when alternatives
provided
29
Question Wording and Framing
Wording of the Response AlternativesSocial desirability
Responses swayed toward socially desirableTerm strength
“Not allow” vs. “forbid” alter responses
30
Question Wording and Framing
FramingDecision maker’s conception of the acts,
outcomes, and contingencies of a choiceFormulation of the problemNorms, habits, and characteristics of decision
maker
31
Question Wording and Framing
Framing Gains vs. losses
Risk adverse when gains are at stake Risk seeking when losses are at stake
32
Question Wording and Framing
Psychological accounting Framed as direct consequence of act or
evaluated with respect to previous balance Paying $10 and losing $10 vs. paying $20
33
How Questions Affect Answers
Question wording and framing influence responses significantly Therefore, be aware of causes and effects
34
How Questions Affect Answers
1. Order of questions
2. Context in which questions appear
3. Format: open-ended or closed
4. Use of filters
5. Use of “catch phrases”
6. Range and order of response alternatives
7. Use of middle categories
8. Framing in terms of gains and losses
35
How Questions Affect Answers
Relevance As a surveyor
To correct for these factors, elicit responses in a variety of ways and compare the results
As a respondent Be aware of the use of techniques designed to
influence your responses
36
SECTION 3
MODELS
OF
DECISION MAKING
3737
Models of Decision Making
Expected Utility TheoryParadoxes in RationalityDescriptive Models of Decision
Making
38
Expected Utility Theory
Nicholas Bernoulli “St. Petersburg Paradox”Two Rules
1. An unbiased coin is tossed until it lands on Tails
2. The player is paid $2 if Tails comes up on the opening toss and doubled for each subsequent Tails
How much would you pay?
38
39
Expected Utility Theory
39
Why the “Paradox”?
(INFINITE)
40
Expected Utility Theory
Daniel BernoulliNicolas’ cousin solved :
40
“Utility” or value declines with amount won
4141
Expected Utility Theory
Declining Marginal Utility
Actual Value$1 + $.75 + $.50 + $.30 + $.15 + $.5 ≈ $2.75
Expected Value$1 + $1 + $1 + $1 + $1 + $1… =
4242
Oskar Morgenstern Jon von Neumann
Expected Utility Theory
Classical utility theory doesn’t describe how
people actually behave
Normative Model
It describes how people behave if they follow certain
testable assumptions
(1947)
4343
Expected Utility Theory
Maximizing Expected Utility
1. Ordering of alternatives
2. Dominance
3. Cancellation
4. Transitivity
5. Continuity
6. Invariance
4444
Expected Utility Theory
Expected Utility Family
Stochastic models of choiceDuncan Luce (1959)
Subjective Expected Utility TheoryLeonard Savage (1959)
4545
Paradoxes in Rationality
Cancellation Principle ViolatedMaurice “Allais Paradox” (1959)
Alternative A: $1,000,000 for sure
Alternative B: A 10 percent change of getting $2,500,000, an 89 percent chance of getting $1,000,000, and a 1 percent chance of getting $0
Alternative A: An 11 percent chance of getting $1,000,000 and an 89 percent chance of getting $0
Alternative B: A 10 percent chance of getting $2,5000,000, and a 90 percent chance of getting $0
4646
Paradoxes in Rationality
Why the “Paradox”
After Cancellation:
Both choices offer identical alternativesAlternative A: A 11 percent chance of getting $1,000,000, and a 89 percent
chance of getting $0
Alternative B: A 10 percent change of getting $2,500,000, and a 90 percent chance of getting $0
4747
Paradoxes in Rationality
Transitivity Principle Violated
Remember:If A > B and B > C then A > C
However, if there are:1. Two dimensions
2. Increase in small inversely related steps
4848
Paradoxes in Rationality
Intransitivity Demonstrated
1. If IQ different by more than 10 pick higher IQ
2. If not pick most experience
DIMENSIONS
Intelligence (IQ) Experience, Years
A 120 1
APPLICANTS B 110 2
C 100 3
B > A and C > B but A > C
4949
Paradoxes in Rationality
Do people violate Transitivity?
Amos Tversky (1969)
Short answer: YES!
Gamble Probability of a Win
Payoff, $ EV, $
A 7/24 5.00 1.46
B 8/24 4.75 1.58
C 9/24 4.50 1.69
D 10/24 4.25 1.77
E 11/24 4.00 1.83
6 out of 18 Harvard Grad Students demonstrated reliable intransitivity
5050
Paradoxes in Rationality
Are Violators Irrational?
Approximations provide a
rational decision strategy
“Expected Utility Theory does not adequately describe how people make
decisions”.
51
Descriptive Models of Decision Making
Normative vs. Descriptive Models
Normative – how rational actors behave when assumptions are met
Descriptive – how people actually behave
51
52
Descriptive Models of Decision Making
Prospect Theory
52
•“Utility” replaced by “Value”
•Loss aversion leading to “endowment effect”
•Probabilities not perfect and based on decision weights
•Reference point based
53
Descriptive Models of Decision Making
The Certainty Effect
Probabilities based on decision weightsElimination of risk weighed more than a
reductionPseudocertainty
53
54
Descriptive Models of Decision Making
Multi-Attribute ChoiceCompensatoryTrades off low values on one dimension
against high values in another
Noncompensatory1. Conjunctive Rule2. Disjunctive Rule3. Lexicographic Strategy4. Elimination-by-aspects
54
55
Descriptive Models of Decision Making
55
SatisficingHerbert Simon (1956)
•People “satisfice” not optimize
Alternative Theories
Regret TheoryBell, Looms & Sugden
•Faced with uncertainty people will act in anticipation and avoidance of regret
56
Descriptive Models of Decision Making
ConclusionSt. Petersburg Paradox led to Expected
Utility TheoryProblems such as Allais Paradox led to
alternativesProspect theory a descriptive model is
now widely accepted and applicable
56
57
SECTION 4
HEURISTICS
AND
BIASES
58
Heuristics and Biases
Complicated judgments or decisions are simplified by general rules of thumbClose approximations to optimal
answersCreate predictable biases and
inconsistencies
59
Heuristics and Biases
Representativeness HeuristicAvailability HeuristicProbability and RiskAnchoring and AdjustmentPerception of RandomnessCorrelation, Causation and ControlAttribution Theory
60
The Representativeness Heuristic
People judge probabilities based on the degree to which “A” is representative of “B”Person “A” is representative of
group “B”Event “A” was produced by
process “B”
61
The Representativeness Heuristic
Conjunction FallacyThe Law of Small NumbersNeglecting Base RatesNonregressive Prediction
XOXXXOOOOXOXX
62
The Representativeness Heuristic
Actuarial PredictionsDo not be Misled by Detailed
ScenariosPay Attention to Base RatesRemember Chance is not Self-
CorrectingDo not Misinterpret Regression
to the Mean
63
The Availability Heuristic
People assess the frequency of a class or probability of an event based on the ease with which instances or occurrences can be brought to mind
64
The Availability Heuristic
Availability Can Go AwryImagination MAY Affect AvailabilityVividness Discussed
65
The Availability Heuristic
Availability heuristic is often OK Can also lead to critical biases Central to frequency and probability
estimates Do not trust your immediate
intuitions for judgments of frequency or probability
Consider the cause of your intuition Test your intuition
66
Probability and Risk
Risk is omnipresent and often difficult to quantify
Basic problems in probabilities can be challenging
67
Probability and Risk
Confusion of the InverseValence EffectsCompound EventsConservatismThe Perception of Risk
68
Probability and Risk
Maintain Accurate Records Beware of Wishful Thinking Remember Prior Probability Break Compound Events into Simple Events Beware Risk Perceptions
69
Anchoring and Adjustment
Insufficient adjustment up or down from an original starting value – the anchorRobust phenomenonAffects most people across
many circumstances
70
Anchoring and Adjustment
Staking out initial positionsFramingImagination
71
Anchoring and Adjustment
Form judgment without the frame first
Check if starting value is diagnostic Beware unusually high or low values Use imagination / reflection to
counteract the affect of anchoring Be aware of best / worst case
scenarios
72
The Perception of Randomness
Human often misperceive and misconceive random events
73
The Perception of Randomness
Coincidences are common, but particular coincidences are not
People often see meaning / patterns in random events
Superstitions Behaving randomly
74
The Perception of Randomness
Do not over-interpret chance eventsIf there are independent events with
equally likely outcomes, do not view short runs of the same outcome as meaningful
75
Correlation, Causation, and Control
It is often difficult for people to see correlation
Correlation between two variables does not imply there is a cause-and-effect relationship between the two
76
Correlation, Causation, and Control
People generally have no adequate concept of correlation
Illusory correlationsInvisible correlationsCausalationControl
77
Correlation, Causation, and Control
Focus on more than positive and confirming cases of relationship
What did not take place may be significant
Determine if judgments rest on observations or expectations
Carefully distinguish between correlation and covariation
78
Attribution Theory
Theory about how people make “casual” attributions or explanations for the causes of actions and outcomes The Person (Consensus) The Entity (Distinctiveness) The Time (Consistency)
“So, if it’s good it’ Mister Coffee.
If it’s bad, it’s me.”
79
Attribution Theory
Ignoring the Base Rate of ConsensusSalienceThe Fundamental Attribution ErrorSelf-serving BiasEgocentric BiasesPositivity EffectAscribing Variability
“If the jury had been sequestered in a nicer hotel, this would probably never have happened.”
80
Attribution Theory
Pay close attention to consensus information
Put yourself in others’ shoes
Look for hidden causes (beyond salience)
Be careful of wording and phrasing
81
SECTION 5
THE SOCIAL SCIENCE OF JUDGMENT AND
DECISION MAKING
82
Section 5 – The Social Side of Judgment and Decision Making
Social Influences
Group Judgments and Decisions
83
Social Influences
“People are social by nature, so their judgments and decisions are subject to social influences”
Social factors’ pivotal role on decisions/actions “How will others react if I blow this
presentation?”
84
Social Influences
Social Facilitation The presence of audience
affects the performance
Social Loafing & Bystanders “None of my
business…”
85
Social Influences
Social Comparison “People evaluate their opinions and abilities
by comparing themselves with similar others.”
“Social comparisons can influence perceptions” “Social Analgesia”
ConformityPeople conform to the majority view, even if it
contradicts to their own judgment.
86
Social Influences
Groupthink In-group pressure to deteriorate judgment
87
Social Influences
Groupthink Symptoms
One man to stand out – but WHO???
- Over optimism and excessive risk taking- Efforts to rationalize or discount warnings- Unquestioned belief in group’s morality- Seeing adversaries too evil, weak or stupid
- Pressure on dissenters from the majority view- Illusion of unanimity- Self censorship of deviations from the apparent group consensus- “Mindguards” who stop controversial information
88
Social Influences
HENRY FONDA!
89
Social Influences
Impacts of social factors should be taken into consideration for decisions/judgment/performance
Diffusion of responsibility
Factors could influence decisions with or without underlying judgments
90
Group Judgments and Decisions
“Do groups make better decisions than individuals?”
“Do groups operate with the same heuristics and biases as individuals?”
91
Group Judgments and Decisions
Group Errors and BiasesGroup attribution errorGroup serving biasesOutgroup homogeneity bias
92
Group Judgments and Decisions
Group Polarization “Group discussion to
amplify the inclinations of a group”
“Risky shift”
93
Group Judgments and Decisions
Group discussion Better decisions? “Groups usually perform somewhat better than
average individuals” Dictatorship Beneficial???
94
Group Judgments and Decisions
“Groups usually outperform individuals”
“The best member often outperforms the group”
“However, the best is not to have more heads, but to put those heads together”
95
SECTION 6
COMMON TRAPS
96
Overconfidence
Disasters caused by overconfidencePearl Harbor in World War IIThe U.S. space shuttle ChallengerOthers…
97
Overconfidence
Joseph Kidd ExperimentA case study of “Joseph Kidd” includes four
partsAsk psychologists, grad students, and under
graduates to answer the same questions when they finish each part of the case study
Ask them to rate confidence level
98
Overconfidence
Joseph Kidd Experiment Results
99
Overconfidence
Joseph Kidd Findings No significant differences
among ratings from psychologists, graduate students and undergraduates.
Confidence increased with the amount of information subjects read, but accuracy did not.
100
Overconfidence
Lichtenstein and Fischhoff Overconfidence is greatest when
accuracy is near chance level. The gap between accuracy and
confidence is smallest when accuracy is around 80%
It is not related to a decision maker’s intelligence
101
Overconfidence
Overcoming overconfidence May want to flag certain judgments for special consideration.
Overconfidence is greatest when judgments are difficult or confidence is extreme. In such cases, it pays to proceed cautiously.
May want to "recalibrate" your confidence judgments and the judgments of others. If a decision maker is 90 percent confident but only 70 to 75 percent accurate, it is probably best to treat "90 percent confidence" as though it were "70 to 75 percent confidence."
May want to automatically convert judgments of "100 percent confidence" to a lesser degree of confidence. One hundred percent confidence is especially unwarranted when predicting how people will behave.
102
Self-Fulfilling Prophecies
“If a card has a vowel on side, then it has an even number on the other side.”
Which of the cards would you need to turn over in order to decide whether the statement is true?
E K 4 7
103
Self-Fulfilling Prophecies
Most common response was “E and 4”
“E” was the next most common Only five of 128 gave the correct
answer: “E and 7”
104
Self-Fulfilling Prophecies
How to reduce self-fulfilling propheciesFocus on motivational factorsFrame questions in a way that
encourages disconfirming answersConsider why their judgments might
be wrong
105
Behavioral Traps
A behavioral trap is a situation in which individuals or groups embark on a promising course of action that later becomes undesirable and difficult to escape from
106
Behavioral Traps
Time Traps Momentary gratification clashes with long-term consequences
Ignorance Traps Unknown consequences
Investment Traps Prior expenditure leads to sunk costs
Deterioration traps Similar to Investment traps, with changing costs
Collective Traps Involve more than one party, such as prisoners’ dilemma and
the tragedy of the commons
107
Behavioral Traps
How to escapeConsider the costs of withdrawal
before embarking on a long term venture
Set limits in advance whenever possible
Have different people make initial and subsequent decisions
108
Behavioral Traps
Traps are not always bad.Recovering alcoholicsEx-smokersOthers….
109
Conclusion
This book shows how a social perspective on judgment and decision making can offer practical suggestions on how to deal with many common problems in daily life. It helps us to understand human thought, feeling, and behavior. It also can help us to avoid decision biases, errors, and traps
Western bias in the judgment and decision making research
Judgment and decision making research can be fraught with biases itself…