MAKING DECISIONS CLPS 0220
adding options
- adding options is like adding time => more time you give someone to do a task, the less likely it is for them to get it done - account for procrastination
the deliberative system (dual sys. of learning)
- analytic: abstraction plus rule application - agents assumed to be conscious of deliberative processing and this enables rule following -requires effort -takes time -may be related to "secondary" emotions: regret, frustration, etc.
monty-hall (three doors) problem
- majority of people will not switch doors when a goat is revealed behind one door - in reality the probability of getting the prize increases if they do switch
satisficers
- stop searching when main criteria are satisfied -happier
choice and satisfaction trends
-as options increase, selection decreases -as options increase, satisfaction decreases
maximizers
-attempt to find best possible option -tend to search larger option set -process can be open ended -tend to be less happy and have more regret
risk neutral
-dont care about the outcome certainty, just the the utility of the outcome - U(medium) = 0.5U(small) + 0.5U(large)
paradox fo choice
-freedom of choice=>tyranny of choice
the sure thing principle
-if X is preferred to Y knowing that event A is obtained, and if X is preferred to Y knowing that A did not obtain , then X should be preferred to Y even when its not known whether A is obtained -e.g. if you're login to drink red wine when it is raining and you are going to drink it when it isn't raining, then there is no reason to check if its raining outside
power of a story
-narrative structure is persuasive -provide reasons
intuition (dual sys. of reasoning)
-non-analytic -agents are conscious of intuitive responses but not intuitive processing and not rule following -automatic -takes less time -may be related to 'primary'/reactive emotions: disgust, joy, outrage=> "gut feeling"
association availability
-people asses the frequency of a class or the probability of an event by the ease at which instances or occurrences can be brought to mind - we tend to rely on information that is readily available in memory but fail to discount value of that information -recent events are judged to be more frequent - high;y imaginable events judged to be more probable - salient objects are judged to be more frequent -easy to pronounce stock yield more profit
compromise effect
-people like the one in the middle -use extremes and people will use those as a reference point
transaction utility (utility of shopping)
-perceived value of the deal -difference between reference point and the amount paid
asymmetric dominance
-presence of dominant alternative gives people a reason to chose one over the other -whether there is relation of dominance, that dominance gives us a reason -e.g. microwave example
jam study
-set up: customers allowed to sample a variety of jams and given a a dollar off if they purchase a jar -test: 2 groups: low choice group: 6 high choice group: 24 -results: large group brought in bigger crowd but approx. same # of jars were purchased as small group
reinforcement learning
-use trial-and-error to learn about the world, then use what was learned to help guide preferences and decisions
framing (continued)
-we tend to present things in the way that we wish people to behave -this creates a huge moral issue e.g. radiation therapist gives mortality frame & surgeon gives survival frame
utilities of shopping
1. acquisition utility 2. transaction utility
satisfying (SAT)
1. alternative considered one at a time, in the order they occur 2. compare the values pf each attribute of the alternative to its cutoff level 3. if any attribute is below cutoff, reject option 4. choose first option whose attributes satisfy cutoffs 5. if no option passes, relax cutoffs and repeat
how can hindsight bias be reduced?
1. asking participants to explain the outcome that didn't happen 2. generating supporting reasons has little effect on probability judgement, generating contradicting reasons has pronounced negative effects
problems with gambler's fallacy
1. chance is not self-correcting, random events have no memory 2. we impose causality on correlation 3. we see structure in the absence of correlation
explanations for preference reversals
1. choice: high probability of winning used as justification 2. compatibility hypothesis: weight of an attribute enhanced by its compatibility with price
attributes of choice strategies
1. compensatory vs non-compensatory -trade-off attributes or not -require effort and cause conflict (compensatory) 2. selective vs non-selective information processing 3. quantitative vs qualitative reasoning -matter of degree
the value function
1. defined as losses and gains 2. concave gains; convex losses 3. steeper for losses than gains
lexiographic ordering (LEX)
1. determine most important attribute and examine values of all alternative on that attribute 2. alternatives with the best value on the attribute chosen 3. of one alternative is remaining, you're done, otherwise determine second most important attribute and go from there
implications of loss aversion
1. easier to accept a surcharge than forgo a discount e.g. taking money from teachers with poor student performance more effective than giving them money for improvement of student performance 2. size of loss aversion dependent on what you're talking about
reason based choice
1. enriched options 2. impoverish options -appeal to reason not utility or any other theories -has to be justifiable to ourselves and others
examples of normative theories
1. expected utility theory 2.
physicians & most people vs weather forecasters
1. experience 2. repetitive 3. feedback
why do superstitions arise?
1. filling in waiting time 2. script to prepare mind and body for action 3. belief in hidden agent controlling affairs 4. need to explain events
prospect theory
1. framing and editing 2. evaluation
deferring choice because of conflict
1. low conflict - easier for people to make decisions 2. high conflict -harder for people to make decisions and will delay to make decision easier later on
How do we judge probability?
1. mental stimulation 2. cold calculation 3. rough calculation 4. heuristics
preference reveals understanding utility
1. people know their values directly - e.g. multiplication table -e.g. " I cal them as they are" 2. values exist. they are perceived an reported as best hey can be -e.g. body temperature -e.g. " I call them as I see them" 3. values constructed during process of elicitation. preference is constructive, context dependent process -"they aint nothing till i call them"
weighted additive rule (WADD)
1. put each tribute on a common scale 2. weight each attribute according to importance and how it trades off with the other alternatives 3. for each option, multiply each value by weight 4. utility of option set to sum 5. must assume independence 6. not descriptive
more choices => miserable
1. real: opportunity cost 2. psychological: regret and anticipated regret 3. escalation of expectations (more choice=> we expect more)
3 types of risk utility curves
1. risk aversion 2. risk seeking 3. risk neutrality
eliminating-by-aspects (EBA)
1. same as LEX but most important attribute chosen probabilistically 2. all alternatives with values below the cutoff on that attribute is eliminated 3. if more than on alternative remains, choose next important attribute
equal weight heuristic (EQW)
1. same as WADD but each attribute given equal weight 2. requires a common scale for all attribute values -e.g. salary=1, time=1, prestige=1
principles of hedonic framing
1. segregate gains -segregate small gains from larger losses 2. integrate losses - integrate smaller losses with larger gains
reasons for overconfidence
1. tendency to focus on supporting reasons 2. ambiguity aversion
dual systems of reasoning
1. the deliberative system 2. intuition
frequency of bad and good features (FRQ)
1. use cutoffs to separate good from bad values on each attribute 2. count # of good and bad attributes for each alternative 3. choose alternative with most good attributes
components of decisions
1. value (preferences) 2. uncertainty (belief)
case-study approach
Ben Franklin -advantage: easy to explain and justify; easier to identify what is important to you -disadvantage: do people have access to their own reasons?
bayes rule
P(E1 | E2) = P(E2 | E1) x P(E1) / P(E2)
risk aversion
CONCAVE - made in the domain of gains -comparing and outcome (medium) to an uncertain outcome (small vs large) - U(medium) > 0.5U(small) + 0.5U(large)
risk seeking
CONVEX - made in the domain of losses - puts more utility on large outcomes - winning medium doesn't mean s much - U(medium) < 0.5U(small) + 0.5U(large)
independence
P(A and B) = P(A) x P(B)
conjunction rule
P(A) > P(A and B)
conditional probability
P(A|B) = P(A and B) / P(B)
sample space
a collection of primitive events such that on of them must occur
categorical imperative (immanuel kent)
act only according to that maxim whereby you can, at the same time, will that is should become a universal law
probability
assignment of numbers between [0,1] to propositions representing degree of uncertainty
Pascal's theory
because the rewards and costs are so high for believing in God, it does not matter what the probability of God actual existing is => you should believe in God
superstition
belief that two independent events are dependent despite knowing they are independent
conjunction fallacy
cases where people's probability estimates violate the conjunction rule
habit heuristic
choose what you chose last time
conjunction fallacy
choosing P(Linda is a feminist and a bank teller) > P(Linda is a bank teller)
decision
choosing to do one thing from a set of options were you could have done something else
paradox of choice
conventional wisdom deliberation > intuition
sunk cost
cost that has already been incurred and cannot be recovered
moral dumbfounding (Hadit)
dogmatic insistence on a moral judgement for which no good reasons can be given -e.g. brother and sister traveling together
hedonic framing
evaluating joint outcomes to maximize utility
impoverished option (reason based choice)
has average attributes
enriched option (reason based choice)
has good and bad attributes
qualities of a nudge
libertarian paternalistic manipulates architecture of choice to create desirable outcomes
descriptive theory
how people actually behave
normative theory
how people ought to behave
recognition heuristic
if one of two objects is recognized and the other is not then infer that the recognized object has higher value with respect to criterion
disjunction effect
immediate decisions prevented by an uncertain, but irrelevant, future event
hindsight bias
inclination, after event has occurred, that the event was predictable despite having no or little objective bias for predicting it
acquisition utility (utility of shopping)
measure the value of a good obtained relative to its price
rational analysis of choice
more choice should be good for a rational agent ad they increase the chances of getting greater utility
payment depreciation
more you pay for something the more you will wear them
human sentalism
mortal or ethical sentiments are intrinsically motivating
failed nudge example
nudge: making parents pay every minute they are late to pick up kids from daycare goal: trying to lower the number of kids getting picked up late result: more parents showed up late because guilt they felt by showing us late was dissolved by their payment
endowment effect
owning something gives it more worth
payment decoupling
payment is separated rom consumption so perceived cost of activity is reduced
ambiguity aversion
people like to bet in domains where they are confident and if they are not, they will leave it up to chance
judgment by stimulation
people often judge events they can mentally "simulate" to be more likely
anchoring and adjustment
people often make insufficient estimates by starting at an initial value an adjusting to yield a final answer
planning fallacy
people overwhelmingly underestimate time and costs of their own projects
proximity heuristic
people use judgements of closeness (distance) to estimate risks and probabilities
greatest happiness principle (John Stewart Mill)
precursor to expected utility theory
libertarian
preserve freedom of choice
representativeness
probability that object A belongs to class B originates from process B, is evaluated by degrees to which A resembles B
topical account
relates consequences of possible choices to a reference level that is determined by context
veil of ignorance (john rawls)
resources, positions, and rights should be distributed in society as if you don't know your status, capacities, or interests
paternalistic
steer people's choices in directions that promote their own welfare
proposition
string of symbols that is either true or false e.g. it will rain tomorrow e.g. this coin will come up heads
events
subset of a sample space
gambler's fallacy
the belief that a particular outcome is due if it hasn't occurred in a while
mental accounting
the implicit methods individuals use to code and evaluate financial outcomes
framing effects
the way in which an event, problem or question is worded to produce swayed results
lesson of expectations
what matters to satisfaction is whether experience meets expectations or not
rationality
when descriptive and normative theories converge
money pump
when you violate the transitivity axiom (don't follow expected utility theory)