Reasoning

Decision making under risk involves using

the value and probability of various outcomes

Heuristic

cognitive shortcuts

Heuristic facts (1) useful because of

time/resource limitations

Heuristic facts (2) not guaranteed but

works well a decent amount of the time for it to be useful

The use of heuristics leads to

judgement biases

Anchoring

Strategy in which estimation begins with an initial anchor and the estimate is then adjusted in light of incoming info

Anchoring and adjustment in real world implications

in bargaining price is the anchor, in legal contexts requested compensation is anchor

Almost anything can serve as an anchor but

we often aren’t aware of it

Anchoring example

social security number and code number experiment

Heuristics representativeness

estimate something that is more probable if it has many features you associate with a particular outcome

Typical outcome

people rate the probability of the first scenario higher because of representativeness

Conjunction fallacy

two conditions can NEVER be more probable than the probability of one of them alone

Conjunction fallacy example (1) people have heart attacks with a certain probability

OF THOSE, a high portion may be over 55 or smoke

Conjunction fallacy example (2) but the probability of a heart attack is

overall still higher because that includes minorities in other categories

We know coin toss outcomes should be random so a mix of

heads and tails seems more representative compared to all tails but they’re equally likely

More representative

seems more probable

Representativeness summary

good representatives of classes seem more likely to belong to that class

Representativeness fails when

focusing on similarity to other class members distracts from how frequent that class actually is

Availability

items that are easier to remember are also thought to be more common

Solo members experiment: Listen to 6 different voices and remember and evaluate the statement made by each voice

Group A: told that 5 voices are white and 1 asian

Group B: told that 3 voices are white and 3 asian

Solo member results

the person labeled asian in groups were remember better by group a than b and more negatively evaluated by group A than B

The Why’s of the results for solo members experiment (1) low

frequency of “asian” individuals in group A makes them more distinctive

The Why’s of the results for solo members experiment (2) therefore,

negative behaviors were also more unlikely and more distinctive

More distinctive =

more available = estimated to be more probable

Pros of availability (1) availability reflects the

effectiveness of our memory search strategy

Pros of availability (2) can arise because of

higher frequency of events

Cons of availability

other things can increase availability: recency, publicity, and distinctiveness

Base rate

how often an event occurs in the population (ex. 70% of cats are tabbies)

Sample size

how many observations are made (ex. a vet treats 20 cats in a day)

Smaller sample are more likely

to deviate from the average

Taxi cab problem (1) out of 100 cabs,

85 are green, of those, witness says 20% (17) are blue

Taxi cab problem (2) 15 are blue,

of those, witness says 80% (12) are blue

Taxi cab problem (3) of times that witness would report a blue cab,

12/(17+12=29) = 41% were actually blue

All tests have some probability of false positives

if the base rate of the actual disease is low, there are likely more false positives than correct diagnoses

Ignoring base rate

overestimating how likely it was that we observed an event

Ignoring sample size

overestimating ow much the actual “population” statistics affects a small sample

Inductive reasoning

takes an example or small set of examples and infers what might be true of the group

Deductive reasoning

takes general principles and determine what must follow them

What’s within deductive reasoning

logic

What’s within logic

premise, conclusion, antecedent, and consequent

Logic

a formal system that gives us rules for determining the truth of conclusions based on a set premises

Premise

something we assume/know to be true

Premise example

if a lecture is posted to the psych 224 moodle, then it’ll be about psychology

Conclusion

something we derive from the premise

Conclusion example

therefore, it’s about psychology

Antecedent

the “if” part of a premise

Antecedent example

if a lecture is posted to the psyc 224 moodle, then it’ll be about psychology

Consequent

the “then” part of a premise

Consequent example

if a lecture is posted to the psych 224 moodle, then it’ll be about psychology

Affirming the antecedent (1) “if this item

is an apple, then it’s a fruit”

AFFIRMING THE ANTECEDENT (2) “this item

is an apple”

Conclude “therefore, this item

is a fruit”

Affirming the

antedecent is valid

DENYING THE CONSEQUENT "this item is

NOT a frfuit”

Conclude “therefore, this item is not

an APPLE”

Denying the

consequent is valid

AFFIRMING THE CONSEQUENT “this item is

a fruit”

Conclude: therefore this item is

an apple”

Affirming the consequent is

NOT valid bc the item could be a pear

DENYING THE ANTECEDENT “this item is

NOT an apple

Conclude “therefore this item is

NOT a fruit”

Denying the antecedent is

NOT valid bc the item could be a pear

Wason Selection Task

logic tells us that affirming the antecedent should be true (check e) and denying the consequent should be true (check 7)

Confirmation bias

the tendency to look for evidence that confirms our hypothesis rather than looking for evidence for falsifying evidence

People and logic (1) people have

difficulty applying rules of logic in everyday reasoning

People and logic (2) in framing effects, people

sometimes apply logic and sometimes not

People and logic (3) instead of reasoning logically

people seem to apply schemas based on their experience

Matching schema shows

a confirmation bias