W12: Judgment, Decisions, and Reasoning

Question 1

Defination of Central idea,
Heuristic and bias for the heuristics and biases program
S4

Kahneman (2011)

Answer 1

1) Central Idea:
Judgment and decision-making often rests on simplifying heuristics instead of extensive algorithmic processing.

2) Heuristic: (from computers)
A simple procedure that helps find adequate, though often imperfect, answers to difficult questions.

3) Bias:
A systematic error of judgment.

Question 2

Forebears for the heuristics and biases program
Herbert Simon 1990
(2)

Answer 2

Bounded rationality:
Humans reason and choose rationally, but only within the constraints imposed by their limited searchand computational capacities.

Satisficing:
“[U]sing experience to construct an expectation of how good a solution we might reasonably achieve, and halting search as soon as a solution is reached that meets the expectation

Question 3

Forebears for the heuristics and biases program:
Paul Meehl(1954):
2 major findings for clinical

Answer 3

1. Clinical prediction performs very poorly relative to statistical prediction.
2. Clinical prediction overweights case characteristics and underweights base rates

Question 4

2. The two systems of cognition

Dual-process theories of higher cognition

Answer 4

System 1:
Intuitive
Fast
Non-conscious
Automatic

System 2:
Reflective
Slow
Conscious
Controlled

characteristic of System 2 is that it requires attention

Question 5

How to improve System 2?

3

Answer 5

1) Reward to motivate participants to check their intuitive impressions.
2) Ensure participants are not simultaneously having to perform other kinds of mental effort.
3) Another idea is commonly cited, but does not work:- Increase metacognitive difficulty (Alter et al, 2007; Meyer et al, 2015)

Question 6

The proposed mechanism of System 1

Answer 6

Question substitution:
seek the answer to an easier question you believe to be related rather than the complex one

Substitute an easy-to-compute feature for a hard-to-compute feature.

Question 7

The three general-purpose heuristics

Answer 7

a. Representativeness heuristic
b. Availability heuristic
c. Affect heuristic
d. ? Anchoring heuristic

Question 8

a. Representativeness heuristic

Answer 8

Probability judgments (the likelihood that X is a Y) 
substituted with assessments of resemblance (the degree to which X ‘looks like’ Y).

Eg.
“How likely is it that Tom is a computer science student?”
substituted with
“How much does Tom resemble a computer science student?”

Question 9

a. Representativeness Heuristic
Biasness
(6)

Answer 9

1) Conjunction fallacy
2) Insensitivity to sample size
3) Misperceptions of randomness
4) Belief in ‘The Law Of Small Numbers’
5) Hot hand ‘fallacy’
6) base rate neglect

Question 10

a. Representativeness heuristic
(Linda problem)
BIAS: 
1) Conjunction fallacy 
Tversky & Kahneman, 1983

Answer 10

Conjunction fallacy occurs when it is assumed that specific conditions are more probable than a single general one.

Question 11

a. Representativeness heuristic
BIAS:
2) Insensitivity to sample size
3) Misperceptions of randomness

Answer 11

2) Insensitivity to sample size
judge the probability of obtaining a sample statistic without respect to the sample size: variation is more likely in smaller samples, but people may not expect this
Hospital problem (smaller should have )

3) Misperceptions of randomness
Apophenia -
Perceiving patterns in random data

Question 12

a. Representativeness heuristic
4) Belief in ‘The Law Of Small Numbers’
5) Hot hand ‘fallacy’

Answer 12

4) Belief in ‘The Law Of Small Numbers’
Putting in too much faith in studies with a small sample size

5) Hot hand ‘fallacy’
people who’ve achieved recent success have a (temporarily) increased propensity to achieve more success.

6) base rate neglect
base-rate information tends to be neglected when specific information about the case is available.

Question 13

b. Availability heuristic and its biases (3)

Answer 13

Factors that come to mind easily are assigned greater weight in the formulation of judgments.
We judge the likelihood/frequency of an event by the ease with which instances of it come to mind.

“What percentage of Hollywood celebrities are divorced?”
substitute:
“How readily do examples of Hollywood divorces come to mind?”

Bias

1) Effectiveness of a search set
2) Egocentric bias
3) Outcome bias

Question 14

b. Availability heuristic

1) Effectiveness of a search set

Answer 14

Its easier to search for something starting with a letter H than having H as the second letter in a word, even though its more common. 1st letter is more availability to us…

Question 15

b. Availability heuristic

2) Egocentric bias

Answer 15

Tendency to rely too heavily on one’s own perspective and/or have a higher opinion of oneself than reality.

Question 16

b. Availability heuristic

3) Outcome bias

Answer 16

A decision with a positive outcome is rated as superior to a decision with a negative outcome even when the information available to the decision maker was the same in both cases.

Question 17

Proposed mechanism of availability

Is availability based on the amount of content we recall, or on the ease with which the content is recalled?

Answer 17

Schwarz et al, 1991:

Task: State either 6/12 examples of either assertive/unassertive behaviors. (4 grps)

dependable:
Self reported assertive level after task.

Determinant is the effortfulness of recall of examples:
6 easier than 12. Thus, higher in rating for the associated group/.

Question 18

C. Affect heuristic and biases

Answer 18

judgments are made in accordance with the intensity of the emotion felt.

“How large are the benefits of nuclear power?” substitute: “How do you feel about nuclear power?”

Bias:

1) Negative risk/benefit correlation
2) Insensitivity to numbers

Question 19

C. Affect heuristic
Bias:
1) Negative risk/benefit correlation

Answer 19

In the real world: risk and benefit positively correlated.

In people’s minds and judgments: risk and benefit negatively correlated.

Finucane et al (2000) found that information about benefits impacted the perception of risk, and vice-versa

Question 20

C. Affect heuristic
Bias:
2) Insensitivity to numbers
Slovic & Peters (2006): airport safety study

Answer 20

question “How much would you support the proposed measure to purchase the new equipment?”

Saving a % of 150 lives was more supported than saving 150 lives

Question 21

d. anchoring heuristic and biases

Answer 21

According to this heuristic, when estimating quantities people start with an intuitive reference point (the ‘anchor’) and make adjustments to it.

Biases:

Question 22

Practical implications of anchoring

Answer 22

Studies have investigated the practical implications of anchoring in various domains of life

• Criminal sentencing
  
  • Ahead of a mock sentencing, judges
    provided random anchors for themselves
    using dice. (Englich et al, 2006)
• Salary negotiations
  
  • Galinsky & Mussweiler (2001)
• Consumer behaviour
  
  • Simonson & Drolet (2004)

Question 23

Three criticisms of K-T’s work

Answer 23

1. K-T’s biases lack external validity
2. K-T’s heuristics are vaguely theorized/specified and there has been a lack of formal modeling
3. K-T overstate the problems caused by the computational limitations of our brains

Question 24

Three criticisms of K-T’s work
1. K-T’s biases lack external validity

Researchers noticed that minor changes of wording/context could sometimes ‘debias’ people:

Answer 24

1. Eliciting frequencies rather than probabilities
2. Emphasizing the role of random sampling

Findings are aimed at questioning the external validity of Heuristics & Biases studies

the criticism suggests that results from the studies might be artefacts of experimental design, and that the results might not generalize beyond the lab.

changing probability to frequency
Emphasizing random sampling -> Responses greatly improved (lawyer/engineer)

Question 25

Three criticisms of K-T’s work

2. K-T’s heuristics are vaguely theorized/specified and there has been a lack of formal modeling

Answer 25

• Problem that heuristics like ‘availability’ and ‘representativeness’ are so general they can explain a wide range of findings post hoc
• Suggestion that formal models of heuristics should be proposed that make quantitative predictions (e.g. about rates of mistake judgment).
• Some promising attempts involve a causal Bayesian framework (e.g. Krynski & Tenenbaum, 2007).

Question 26

Three criticisms of K-T’s work

3. K-T overstate the problems caused by the computational limitations of our brains

Answer 26

• Gigerenzer points out that even if the mind had more computational resources, it does not follow those complex methods will outperform simple ones.
• “Fast and frugal” methods can be better
  Example: The recognition heuristic.
  the recog heuristic can help!

Question 27

PLS LOOK AT SLIDE 73

Answer 27

LOOK AT SLIDE 73