Cognitive psychology Judgement + Decision making

Khanneman & Tversky, 1972

Taxi-cab probability problem, most people get it wrong because they fail to take into consideration base rates (the frequency in which something naturally occurs in a population)

heuristics

simple rules of thumb that result in approximately correct judgements, usually useful, sometimes colored by bias

Tversky & Kahneman (1973) availability heuristic

over-estimating the frequency of events that are readily accessible

2 mechanisms of the availability heuristic

• fluency mechanism: how easily do examples come to mind, biases toward personal experience

• availability-by-recall mechanism: judging frequency by how many relevant instances you can recall (usually influenced by the news)

factors that influence availability heuristic

1. vividness of memory

2. personal experience

3. unusual events

4. media reports

5. primacy of occurence

6. recency of occurence

support theory Tversky & Koehler, 1994

subjective probabilities of events are perceived as higher when they are more explicitly described (sick vs COVID or cold or flu- latter seems more likely)

→ based in the availability heuristic

anchoring heuristic Tversky & Kahnemann (1974) roulette wheel

when estimating probability, people anchor on an implicitly or explicitly suggested reference point

→ roulette wheel rigged to stop high or low, impacted how many UN countries are in Africa - usually used for info we have little internal information about

→ Ghandi age experiment (9 or 150) Strack & Mussweiler (1997)

representativeness heuristic

assumptions based on previous observations and experiences - assumes categories are relatively homogenous, single case representative of class (STEREOTYPES)

conjuntion fallacy: Tversky & Kahnemann,1982 - the Linda problem

ANDS will always make a probability less likely than a blanket statement, even if all of the items fit within the schema

→ directly born from representation heuristic

illusory correlation

erroneous belief that independent factors are related (glasses and intelligence)

recognition heuristic

decision is based on

Khanneman dual process model of judgement

system 1 = fast automatic operations, unconscious (heuristics)

system 2 = effortful, conscious - can modify system 1 judgements

Gigerenzer & Hoffrage (1995) on frequencies

natural frequencies (1 in 1000) are easier for humans to understand than percentages (.1%) because humans were evolved to think in whole numbers

relative vs absolute risk

relative risk: no base rate, ‘20% increase’

absolute risk: includes a base rate ‘5% to 6% increase)

→ relative risk reports can be misleading, need to consider the importance of base rates

Linda problem reformulated into frequentist forms Gigerenzer & Hoffrage (1995)

how many out of 100 people with this description are bank tellers, how many out of 100 are bank tellers and feminists

→ reduced conjunction fallacy errors

Neumann & Morgenstern (1947) utility theory

decisions made based on summation of utility and cost

Expected utility = (probability of a given outcome) x (utility of the outcome)

Tversky & Shafir (1992) coin toss betting

shows people are aversive to loss even when the odds are in their favor

→ potential gains must be, on average, twice the potential loss (prospect theory)

Kahneman & Tversky (1979) prospect theory

people are more sensitive to loss than gain, potential gain must be 2x potential loss for people to make the bet

→ also will put too much weight on low probability events when trying to avoid loss

self esteem in prospect theory

Josephs et al. (1992): people with high self esteem will more often take the high risk high reward path because they can withstand failures whereas low self esteem individuals are threatened by risk

sunk cost effect

people will invest more into a failing endeavor to justify not losing their initial cost

→ a learned behavior from adults who believe they will have to explain their actions

Tversky & Kahneman (1981) - the framing effect

people opt for probabilities that are more positively framed even if the actual chances are the same (saving 40% of people more often taken than letting 60% of people die)

→ another form of loss aversion explained by prospect theory

omission bias

people prefer inaction to action when making risky decisions since people feel more responsible for their actions than their inactions

→ Ritov and Baron (1990) vaccine experiment, people would prefer inaction and higher odds of their child dying naturally than their child dying from vaccine

Wright, 1984: multi-attribute theory

attributes of different decisions are weighted and summed, highest weight is the decision that is made

→ rarely employed: high energy cost and requires people having all of the relevant information

→ assumes UNBOUNDED RATIONALITY

Simon, 1957: theory of bounded reality + satisficing heuristic

in reality, there are bounds to rationality (our attention and access to relevant information)

leads to the use of satisficing heuristic (satisfactory + sufficing): weighing options and then selecting the first one that meets the minimum requirements

unconscious thought theory

simple decisions are best made by conscious thought, complex thoughts can be best made by unconscious thought since it has a greater processing capacity

Dijksterhuis and Nordgren (2006)