Test 3 (decision-making) Flashcards
(18 cards)
required for decision-making
- 2+ options
- ability to have expectations of each choice
- ability to make a value-judgement of the outcome of each choise
normative theories of decision making
assume everyone acts in their best interest to maximize the EV
descriptive theories of decision-making
how people actually do behave
prospect theory & exemplar STUDY
assume people are affected by biases and other inputs (ex. framing effects)
STUDY: decision whether to administer vaccine, different based on framed as risk or reward
guiding principles when evaluating risk
- inclined towards loss-aversion (want to avoid losses more than we seek gains)
- overestimate likelihood of low-probability event and underestimate likelihood of high-probability evet
- base decision of current state and wanting to minimize change from current reality
risk preferences
- High-probability gains: exhibit risk-averse behavior
- Low-probability gains: exhibit exhibit risk-seeking behavior (ex. buying a lottery ticket)
- High-probability losses: risk-seeking behavior (ex. Nanaji keeping Alibaba stock)
- Low-probability losses: risk-averse behavior (ex. buying insurance for rare natural disaster)
primary reinforcers
intrinsic needs (i.e. food, sex)
secondary reinforcers
help to obtain primary reinforcers (ex. food)
positive reinforcers
delivers a reward or positive outcome for following intended outcome (ex. a sticker for completing homework)
negative reinforcers
makes an outcome more likely by stimulating behavior to remove something undesired
(ex. putting on your carseat to stop the beeping)
punishment
delivers averse outcome when don’t seek intended outcome (ex. when don’t turn in grade on time, get a grad deduction)
the role of DA in reward-behavior
released from VTA and SN, projects to NA and stimulates vmPFC (reward-seeking behavior)
STUDY: uncued vs. cued rewards
Da neurons fire when expect reward to happen after cue and decreased firing rate when it doesn’t occur
‘reward prediction error’ signal to dlPFC from dACC when wrong to adjust behavior
how do we evaluate risk?
the dmPFC reflects analysis of risk probabilities and the AI reflects the “emotional” signal of risk aversion
temporal discounting
preferring immediate reward over delayed reward
- top-down regulation required to delay gratification
STUDY: ultimatum game
judge fairness of an offer and decide whether to accept or reject (splitting some amount of money)
- when said computer given offer, when its unfair more likely to accept if a computer
activation on AI when reject (remember activity in “emotional” signal of risk aversion)
Drift diffusion model of decision-making
(for 2-choice decision) noise of up and down until reach decision threshold
STUDY: willingness to pay based on how valuable object perceived as
evaluating value associated with vmPFC activation