Dual process theory

Question 1

What is the Cognitive Reflection Test (CRT)?

Answer 1

A series of questions which have intuitive incorrect answers and easily calculable correct ones, often used in job interviews etc. as they’re diagnostic of impulsivity (DRD) and risk seeking behaviour.

Question 2

What is the critical point about the CRT according to Frederick (2005)?

Answer 2

The problems aren’t hard, but they test who’s bothered to check their initial intuitive answer.

Question 3

What is dual process theory?

Answer 3

A common type of theory in cognitive psychology, proposes two systems (Kahneman, 2011) which help us to make decisions.

Question 4

Describe system 1 of the dual process theory.

Answer 4

Heuristic, fast, implicit, automatic, visceral, associative, often used.

Question 5

Describe system 2 of the dual process theory.

Answer 5

Analytic, slow, explicit, deliberative, cold, rule-based, rarely used, monitors system 1.

Question 6

According to dual process theory, what is the role of system 2?

Answer 6

It distinguishes us from animals, who learn associatively, and enables us to do maths.

Question 7

According to dual process theory, what determines people’s decisions?

Answer 7

The relative dominance of the systems.

Question 8

Which system do we use more for ourselves, and which for others?

Answer 8

System 1 for ourselves and system 2 for others.

Question 9

What are the different names for the two systems?

Answer 9

Various - depends on whose model it is (e.g. associative/rule-based, heuristic/analytic).

Question 10

How can conflict monitoring be used to study dual process theory?

Answer 10

When the two systems are placed in conflict, two answers should be produced, which should delay answering in incongruent conditions.

Question 11

What did De Neys & Glumeric (2008) do?

Answer 11

Conflict monitoring with base rate problems, using a think-aloud procedure that manipulated statistical base rates with stereotypical expectations of the representativeness heuristic. System 1 answers would neglect the base rate, system 2 answers would incorporate it. Two conditions - incongruent or congruent (high probability base rate or low, either congruent or incongruent with the representativeness heuristic).

Question 12

What did De Neys & Glumeric (2008) find?

Answer 12

Incongruent: base rate mentioning and accuracy around 20%
Congruent: base rate is just below 20%, accuracy is just below 100%
Neutral: base rate is about 55%, accuracy is around 80%

Question 13

What does attribute substitution do according to Kahneman & Frederick (2005)?

Answer 13

Provide evidence for dual process theory.

Question 14

When does attribute substitution occur?

Answer 14

When people are confronted with a difficult question and try to answer an easier one instead (they’re often unaware they’re doing this) - the harder target attribute is substituted with an easier heuristic attribute.

Question 15

Give an example of attribute substitution.

Answer 15

When asked what proportion of long-distance relationships fail within a year people tend to answer as if they’d been asked “what instances of failed long-distance relationships come to mind”, i.e. the availability heuristic is used.

Question 16

What did Strack, Martin and Schwarz (1988) do?

Answer 16

They asked participants two questions:

1. How happy are you with your life in general?
2. How many dates did you have last month?

Question 17

What did Strack, Martin and Schwarz (1988) find?

Answer 17

The correlation between ratings made on the two answers was small when asked in that order, but when the order was reversed the correlation rose to .66.

Question 18

What do Strack, Martin and Schwarz (1988)’s findings suggest?

Answer 18

That the dating question became the heuristic attribute.

Question 19

How do risks and benefits relate in the environment and analytically?

Answer 19

Although they’re separate dimensions, they have a tendency to be positively correlated. Although activities that are highly beneficial may be high or low in risk, activities that are high in risk but low in benefit are rare and are frequently proscribed

Question 20

How do people behave as if risks and benefits are correlated in the environment?

Answer 20

As if they’re negatively correlated, perhaps because risks and benefits are combined into a single affective utility dimension for heuristic decisions.

Question 21

How does risk and benefit correlation relate to dual processing theory?

Answer 21

System 2 is relatively slow, meaning its operations can be disrupted by time pressure. Therefore the negative correlation in choices in which risk and affect can be obfuscated should be more pronounced when system 2 is impaired.

Question 22

What did Finucane et al. (2000) do?

Answer 22

Tested negative correlation in choices in which risk and affect can be obfuscated - should be more pronounced when system 2 is impaired, so they asked participants to rate the risks and benefits (separately) of various products and technologies under no time pressure or a 5-second deadline.

Question 23

What did Finucane et al. (2000) find?

Answer 23

When participants were required to respond under time pressure their ratings were more negatively correlated, suggesting that both responses were made using the same affective heuristic scale. Under no time pressure the ratings were non-significantly negatively correlated, suggesting that the ratings were independent.

Question 24

How do moral judgements, which are often affective and intuitive in nature, provide a test for the dual process theory?

Answer 24

They might dissociate systems 1 and 2 thinking.

Question 25

What did Greene et al. (2001) do?

Answer 25

Presented people with personal and impersonal versions of the trolley dilemma, in which utility maximisation (saving most people) conflicts with personal morality (killing a person). The two options are either personal or impersonal (pushing someone/flicking a switch).

Question 26

What did Greene et al. (2001) find?

Answer 26

Participants who made the ‘appropriate’, ‘rational’ response took longer to decide in the personal version, which is indicative of a slowness to override the initial heuristic response with an analytic one, suggesting the involvement of systems 1 and 2.

Question 27

What did Greene et al. (2001, 2004) also observe?

Answer 27

Significant differences in neural activity between conditions:
Impersonal: activation in areas of working memory (system 2)
Personal: activation in areas of emotion (system 1)

Question 28

Why are dual process theories popular?

Answer 28

Because the added dimension of freedom appears to give them more explanatory power.

Question 29

What criteria did Osman (2004) apply to test whether key paradigms in thinking were best explained by a dual process theory?

Answer 29

1. Criterion S (Sloman, 1996)
   - Situations in which individuals are led to respond in a manner consistent with system 1 but generate a different (conflicting) response with system 2
2. Individual differences (Stanovich & West, 2000)
   - Large differences are only found when both system 1 and 2 are engaged and produce opposite responses
   - Intelligent people should be more likely to give a system 2 response.
   + Moderate correlation between IQ and performance on any cognitive test.
3. Implicit vs. explicit processes
4. Neuroanatomical differences

Question 30

How does the conjunction fallacy satisfy criterion S according to Sloman (1996)?

Answer 30

Because even 85% of doctoral students in decision science who knew the objective probabilities still commit the conjunction fallacy.

Question 31

What did Stanovich & West (1998) do and find?

Answer 31

Gave participants hard and easy versions of the conjunction fallacy (Linda problem and job problem), found that the mean SAT score for people not committing the fallacy was higher than those who didn’t but effect was larger for the Linda problem compared with the job problem.

Question 32

How did Stanovich & West (1998) justify their finding that the mean SAT score for people not committing the fallacy was higher than those who didn’t?

Answer 32

They argued that this was because overriding the initial system 1 response requires greater supervision of system 2 and therefore correct performance appears to correlate with intelligence.