B1 L4 - Advanced decision making

motor skills

ability to solve a motor problem correctly, quickly, rationally, and resourcefully; thus motor acts aren’t carried out in isolation from cognitive mechanisms, many motor deficits and syndrome are associated with cognitive problems

free will

making a decision arises endogenously (not in direct response to an external stimulus or cue), there are no externally imposed restrictions or compulsions that directly or immediately control subjects’ initiation and performance of the act

three conditions of making a decision, req cognitive processing involving memory, reward, rules to be enacted, etc.

at least two possible choices, expectations can be predicted about the outcomes, value of outcome can be assessed

normative theories

concerned with optimal decision making (normative=an evaluative standard), we build up concepts of probability and expected values and act rationally, based on the concept of probability and expected value

expected utility

we tend to make decisions on what something is worth to us psychologically rather than pure economic value, utilityu and value are related but not linearly (diminishing marginal utility)

rationality theory

we want to maximize our expected utility, if we were purely rational then people would all make the same decisions but we don’t always do that

framing effect

cognitive bias that occurs when the way info is presented influences how we make decisions, this can lead to different choices even when the options are the same

does utility theory actually explain human behavior?

normative models don’t always represent the real world behavior; people will pay to avoid risk (insurance) but will also buy lottery tickets (more than expected value)

prospect theory (Tversky & kahneman)

predicts what people will do not what they should do

reference dependence

people make decisions based on anticipated gains and losses compared to their current state

probability weighting

probabilities are subjectively assessed, e.g. overestimate chance of winning lottery

mapping loss aversion

risk averse for high probability gains, risk seeking for low probability gains, risk averse for low probability losses (insurance), risk seeking for high probability losses (risky stocks) → people hate losses more than they love gains, take active steps to avoid loss

decision heuristics

may also account for our lack of rationality in certain situations

bounded rationality

allows people to solve problems quickly within limitations, these can be expressed as numerous ‘mental shortcuts’ and cognitie biases which have potential to go wrong

availability heuristic, Tversky & Kahneman

there are ~2 times more words that you will know with K in the third position than words that begin with K but people are more likely to come up with mroe words that start with K → is a bias towards more immediate examples that come to mind

representativeness heuristic

thought to derive from ‘the degree to which an event is similar in essential characteristics to its parent population, and reflects the salient features’; focus on a stereotype or perceived generalization of a given category, irregularity can be part of this category (could lead to incorrect decisions simply due to prejudice)

anchoring and adjustment bias

subsequent value negotions might derive from the anchor (when we make assessments by starting from an implict or explicit initial anchor), adjustemnt is when we make insufficient adjustments away from the anchor

AAA bias, Gandhi death estimates

if given nonsense information (Gandhi died at age 9/age 140), estimates of real age of death afterwards were affected

nash equilibrium

decisions with others, set of strategies such that no player has an incentive to change their strategy gien what the other players are doing

the Prisoner’s dilemma

is it optimal to confess and throw someone else under the bus or remain silent? → can’t gain an advantage by changing their strategy in case the other prisoner snitches, factors in the social context to decision making, sometimes the optimal outcome isn’t the most selfish one

dopamine rewards

primary reinforcers (direct benefits like water, food, pleasure), secondary reinforcers (when reward can be used to gain other rewards)

dopamine aversion

removal or reinforcers, aversive stimuli

dopamine plays an important role in reward prediction

reinforce learned behaviors associated with learning outcomes, neurons fire when outcome is better than expected (pos predicition error) and less when outcome is worse than expected (neg prediciton error) → how people become addicted

mobile gaming

designed to be highly addictive (lots of secondary reinforcers), scarcity and intermittent reinforcers; foster addiction with instant to delayed gratifiication, remedies for loss aversion, sunk cost and endowed progress, reciprocity, envy), loot boxes (essentially gambling)

phasic changes in activity of dopamine neurons signal

changes in information

reward prediction error

difference between the reward that was received compared to what was expected and the change in info about future rewards

tmeporal difference learning

successive states of the wrold are correlated over time

actor-critic learning models

critic evaluates whether rewards are better or worse than expected, actor updates the values of potential courses of behavior other increase the likelihood of future rewards; linked to separate components of the basal ganglia

fictive learning

changes in behavior basedo n what might have been, stock that rose unexpectedly would generate real RPE signal in someone who owned it but fictive RPE signal in someone who was simply keeping track of it

salient events

any stimulus that provides info important for future behavior, pos and neg reinforcers

risk aversion

ambiguity aversion, uncertainty about what outcome will occur

delay discounting

uncertainty about when an outcome will occur

areas associated with risk of decision

DLPFC, DMPFC (state of uncertainty leads to inc activation), PPC, anterior insula (modulation of aversive signals that push behavior away from risky options and toward greater certainty)

temporal discounting

people judge rewards to be less valuable the farther in the future they will occur; hyperbolic discounting when rewards are discounted very steeply over short time intervals but more slowly over long time intervals

system 1 (temporal discounting)

fast, parallel, automatic and context dependent; emotionally guided ddecisons like immediately available reward

system 2 (temporal discounting)

slower, serial, controlled and evidence-based; decisions involving deliberation or cost-benefit analysis

choices that involve an immediate reward invole the

ventral striatum and medial/ventral PFC (consistent with system 1), lateral PFC and other executive function regions active regardless of delay until reward (consistent with system 2)

dual-system

choices of short- or long-term rewards depend on the interaction between two systems (1 and 2), but similar design showed magnitude of activation within reward-related regions tracked the subjective value regardless of a delay

social stimuli reinforce a wide range of behaviors

attractive faces evoke activation in ventral striatum and ventromedial PFC (reward regions)

charitable giving explained by

internal reward signals that reinforce the behavior despite its cost (activation of central striatum), prosocial behavior req social congiiton processes that recognize another individual’s needs as a prerequisite for actions to help that individual

Prisoner’s dilemma, cooperation

evokes activation in the nucleus accumbens and other reward-realted regions, may predict inc reward on future trials through development of trust (caudate activation)

hyperscanning

simultaneous recording of fMRI data from individuals in two scanners playing a game against each other

altruistic punishment

censuring people who violate social norms, unfair offers linked to activation in insular cortex

satisficing

identifying an important feature such as price, selects a few candidate items to examine secondary features, then buys the item that is most acceptable

endowment effect

people req mroe money when they are selling something they own than they’d be willng to spend to buy an identical good

framing effect

decisions inconsistent with this effect inc activation in DMPFC and amygdala

neurorealism

tendency to regard brain measurements associated with conclusiosn about behavior as more real and convincing than they are

neuromarketing

research on neuroscience of decision making for applications to real-world consumers