PSY362 exam 2

momentum effect/hot-hand belief

success breeds success; failure brings more failure

gambler’s fallacy

recent run makes opposite outcome “due” to happen to balance outcomes (despite no underlying processes forcing the correction)

random sequence

well-mixed sequence in which outcomes alternate w no conspicuous patterns

clustering

when outcomes bunch together

law of small numbers

mistaken belief that a small sample should resemble its source almost as faithfully as a large one

small sample doesn’t have much stability

sample size neglect

makes temporary variation look like a stable regularity

regression toward the mean

following attempts are closer to ordinary performance (back to more typical/expected result); esp important when a case is selected bc its result is unusual

regression fallacy

attributing the natural return toward the average to an intervention

creates a self-reinforcing feedback loop

counterfactual

what probably would’ve happened w/out the proposed cause

usually unseen/unknown

sampling process

how the sample was chosen

survivorship bias

studying only the cases that endured a process while overlooking those that did not

MPG illusion

bc MPG doesn’t change in direct proportion to fuel consumption, apparently large gains at the efficient end of the scale can save less gas than much smaller gains at the inefficient end

denominator neglect

focusing on absolute # of occurrences (numerator) while ignoring the total # of possibilities (denominator), leading to distorted probability assessments

description invariance

requires choice to remain stable across equivilant descriptions

if not, the change is informative

broad framing

evaluating a series of related choices together rather than judging each one in isolation

one-way doors

choices that, once crossed, are difficult, expensive, or impossible to undo

ex. accepting a position that requires relocating across the country

two-way doors

choices that allow easy return passage

conditional probabilities

probabilities calculated under a specified condition

ex.: probability that someone is hospitalized, given that they are in a coma = very high

probability that someone is in a coma, given that they’re hospitalized = very low

sensitivity

the proportion of people with the condition who correctly receive a positive test result

formula = true positives / true positives + false negatives

specificity

the proportion of people without the condition who correctly receive a negative test result

formula = true negatives / true negatives + false positives

base rate/prevalence

proportion of comparable people with the condition before the test result is known

frequency tree

method of working through base rate, sensitivity, & specificity

natural frequencies

counts of people within the population

useful when conditional properties become difficult to interpret

positive predictive value

the probability that a person actually has the condition (disease) given that they received a positive test result

formula = true positives / true positives + false positives

base-rate neglect

ignoring the starting proportion (base rate)

bayesian updating

new evidence revises an initial probability rather than replacing it

overdiagnosis

finding is real, but detecting or treating it may create burdens w/out improving the patient’s life

takeaway: info gained through testing must be evaluated as part of a decision process, not treated as automatically benefici

affective forecasting

trying to predict how we will feel after an outcome

emotions can distort judgment when particulary vivid emotionally possibility dominates alternatives that’re more probable but easier to overlook

status quo bias

tendency to favor an existing arrangement simply bc it is already in place

“staying” just appears neutral but is indeed a decision you’re making

action effect

bad outcomes following an action generate more regret than comparable outcomes following inactivity

reinforces status quo bias

anticipated regret

expected regret following a decision/outcome

causal hypothesis

an account of the process that produced the facts

WYSIATI — what you see is all there is

cognitive shortcut where brain makes quick judgments using only facts it has right now, ignoring any missing info

coherence

hypothesis elements supported one another & fit into a unified account

ex. physician needs some account of what is causing a patient’s symptoms

underdetermination

the evidence available so far doesn’t uniquely identify the explanation that produced it

a coherent fit is NOT sufficient

cognitive ease

the feeling that the pieces belong together & the situation makes sense

familiarity makes info easier to process — still not proof tho

diagnosticity

extent to which a piece of evidence favors 1 possibile explanation over its plausable rivals

the quality of a judgment lies in the PROCESS

confirmation bias

tendency to seek, notice, or interpret evidence in ways that support an existing belief

selective attention

in a crowded scene, some actions receive attention while others remain in the background

basking in reflected glory — BIRGing

tendency to associate onself with the success of others

cutting off reflected failure — CORFing

attempt to distance oneself from an unsuccessful group

motivational reasoning

thinking pulled toward a desired conclusion

identity protection

conclusion helps defend a valued affiliation

changing one’s mind feels like surrendering to the other side

biased assimilation

new info is judged partly according to how comfortably it joins the belief already in place

easily accept favorable info but subject unwelcome info to more demanding scrutiny

naive realism

tendency to experience one’s own view as a different reading of reality while treating a conflicting view as the product of ignorance, irrationality, or bias

feels like perception rather than judgment

simulation heuristic

mental shortcut, substitutes ease of constructing a mental sequence for a careful assessment of whether that sequence will actually succeed

detailed plan; involved attribute substition

inside view

when planners concentrate on the project’s particular design, personnel, sequence, & challenges

planning fallacy

tendency to underestimate time, money, & effort future tasks will require — even when comparable tasks have repeatedly failed to follow their plans

optimism bias

broader tendency to expect desireable outcomes to be more likely, & undesireable outcomes less likely, than the evidence warrents —> unrealistic starting point

reference class

broad group of past projects similar enough to provide a meaningful comparison

don’t exclude unsuccessful projects!

outside view

supplied by reference class—> changes burden of proof for adjustment

shows how far projects like it usually stray from path

reference-class forecasting

systematic use of comparable projects

strategic misrepresentation

presents favorable estimate bc it improves the project’a chance of approval, even while suspecting final estimate will be worse

decision quality

concerns the process used to choose

outcome quality

concerns what happens after a decision

hindsight bias

once people know the outcome, they overestimate how predictable it would’ve been—> distorts remembered certainty

hindsight even harder to resist when we’re evaluating a choice we made ourselves

counterfactual thinking

allows either story to be compared w/what actually happened

evaluation across cases can reveal what a single outcome cannot

considering the opposite

before choosing, ask what could make the preferred option disappoint & what could make the alternative succeed

after outcome is known, reverse the exercise (alternate endings restore some of the uncertainty that the actual outcome erased)

risk transfer

paying insurance companies to assume much of the financial risk of an accident

self-insurance

instead of paying premiums to an insurer, they retain the risk, reserved money for accidents, & paid claims from own resources

risk pooling

combining risks so larger group (all those who don’t make claims) can cover the losses of smaller group (those who made a claim)

narrow framing

presenting something with a specific outcome

ex. when buying a new laptop, narrow framing is salesperson presenting just 1 frightening option of what could happen to it (to encourage you to buy warranty)

risk policy

a rule chosen in advance for deciding which risks to transfer to someone else & which to retain

over-insurance

buyer repeatedly pays someone else to assume losses that accumulated savings could have covered more cheaply

self-insurance makes the opposite trade

catastrophic loss

when an unpredictable loss would be too large for a household to absorb

low probability doesn’t make them safe to ignore bc consequences could be ruinous

absorbable loss

loss can be paid from savings or ordinary income w/out causing lasting financial damage

sunk cost

an investment of time or money once spent & unrecoverable

sunk cost fallacy

“we’ve spent too much time/money to stop now” — fallacy = allowing that loss to govern next investment

past remains powerful psychologically

mental accounting

people often organize outcomes into separate psychological accounts associated w/a purchase, project, goal, career, or relationship

escalation of commitment

an initital commitment creates pressure to continue, & continuation strengthens that pressure

fresh-start test

if we inherited this problem today, w/no responsibility for its history, would we invest the next dollar?

no personal need to prove something

stopping rule (not from frugal tree)

safeguard to establish before commitment becomes difficult to reverse

creates deliberate opportunities to reconsider (ex. temporary laws)

unconscious bias

occurs when automatic attitudes, associations, or expectations systematically shape judgment outside a person’s awareness

halo effect

a favorable impression created by 1 characteristic colors judgments of other qualities that should be evaluated separately

algorithmic decision making

using an algorithm like 5pt scales to make decisions easier

can still be influenced by bias — but limits its influences

equal weighting

adding every weighting gives each attribute equal weight

there are other decisions that call for unequal weights — weighting simply means deciding that some attributes should count more

conformity

people often adjust what they say, or choose not to say, to fit what appears to be the group’s consensus

groupthink

result of when pressure for agreement becomes strong enough to discourage criticism + narrow the consideration of alternatives

premortem

imagines the failure has already occurred + asks participants to explain what caused it while the plan can still be changed (after first generating their explanations independently before discussing)

alters the social meaning of criticism — ideantifying risks is the assignment

devil’s advocate

appoint 1 person to challenge the emerging consensus

psychological safety

when team members believe they can admit mistakes, ask questions, & raise concerns w/out humiliation or punishment

strongest protection comes from combining climate w/procedure

superforecasters

prediction success persisted across different questions & successive rounds of the tournament

approached uncertainty systematically, examined competing possibilities, + remained willing to change their minds

decomposition

breaking 1 intimidating question into smaller questions that can be investigated more clearly; located uncertainty inside the larger problem

reveals a chain of events w/multiple prerequisites — each creates a more manageable inquiry

forecast trajectory

the path a prediction follows as evidence accumulates

each revision must be justified by what the new evidence actually changes

resolution

ability to distinguish events with different probabilities

good forecasting requires both calibration & resolution

active open-mindedness

treating contrary evidence as info to investigate rather than an obstacle to defeat

intellectual humility

keeping confidence proportionate to the evidence + allowing revision to count as learning rather than failure

experiencing self

lives through each successive moment — each hour is real while it occurs, whether or not it’ll later become memorable

remembering self

looks back and answers Q: how was the trip?

memory compresses — selects scenes, emphasizes turning points, + constructs an account w/a beginning, development, & end (problem when this is mistaken for a complete record of the experience)

peak-end rule

tendency to build an overall judgment around an experiences most intense moments & its ending

these moments don’t always determine verdict, but can stand in for much longer streak of experience

duration neglect

when judging an extended experience, people may surprisingly give little weight to how long its pleasure or pain lasted

ending can transform what came before, even when not necessarily representative or useful (matters when used to forecast

opt-in rule

defaul = no; need to register to become organ donor

opt-out rule

default = yes; treated as potential donors unless they registered an objection

default

option produed when no one acts

opt-in/out rules assign opposite consequences to inaction

active choice

no one forced to donate; required only to state a preference

asks everyone to make a decision, but can make expressing either preference very easy (little effort)

choice architeecture

all the design elements + how they’re arranged

no environment is completely neutral — arrrangement may help shape the preference it appears merely to record

nudge

a change in choice architecture that predictably influence behavior w/out prohibiting options or substantially changing economic incentives

libertarian paternalism

requires people remain free to choose & supports design choices to help people act consistently w/their own goals, values, and preferences (like a parent knowing what’s best for their child)

sludge

excessive or unjustified friction that makes it harder for people to obtain an outcome they want

ex. repeated cancellation screens designed to discourage customers from leaving)

present bias

extra pull of immediate costs & rewards

ex. when trying to lose weight, comfy bed is more appealing than going to gym early