Judgement Bias 2 (Pre-lecture 20)

Question 1

Is it rational to use heuristics?

Answer 1

“Adaptive” critique of K & T’s approach (Gigerenzer et al, 99):

• heuristics are part of rational thought.
• ecologically-rational thinking that serves our environment.
• underpinned by ‘bounded rationality’ theory and notion of ‘satisficing’ (Simon, 57)- basically making the best of our cognitive and external limitations.

Question 2

Bounded rationality (Simon, 57)

Answer 2

Correct or rational thinking given constraints- people are essentially rational.
Thinking isrational if it achieves personal goals yet doesn’t comply with logic:
- cognitive constraints (eg. limited working memory).
- external constraints (eg. lack of time).
Satisficing: thinking that’s effective in terms of meeting goals but imperfect when evaluated against standards of logic.

Question 3

‘Adaptive’ approach

Answer 3

K & Ts heuristics and biases approach too focused on probabilistic thinking.
Evolutionary perspective (eg. Cosmides & Tooby, 94): frequency thinking ‘adaptive’:
- frequencies more evident in real world.
- hard wired to think of things in terms of their frequency, so rationality being akin to logic largely unattainable.

Question 4

Frequencies vs. probabilities

Answer 4

Gigerenzer et al: cognitive biases ‘disappear’ when probabilistic info presented in frequency format.

Study by Griffin & Buehler (1999): no effect of info presentation format on base-rate data neglect biases.

• looked at planning fallacy (phenomenon about people being over-optimistic about how long tasks will take to complete)- despite previous similar tasks not finishing on time.
• goes against Gigerenzer’s claim- they said biases are still there regardless of the format of info.

Question 5

Fast and frugal heuristics

Answer 5

Gigerenzer et al, 99: cognitive ‘toolbox’:
- contents used to make effective judgements and decisions in natural environments.
- fast and frugal heuristics (eg. recognition) good for dealing with uncertain situations- rarely lead to bias (unlike K & T); severe everyday needs and goals.
Evidence often from computer modelling of environmental cues.

Question 6

Gigerenzer et al’s approach

Answer 6

Judgement decision making in everyday life is often undertaking uncertain circumstances- not always error-prone despite cognitive limitations.
Correct/rational judgements and decisions often made under uncertain environments.
- eg. fire fighters working under extreme time pressure to quell blaze- don’t have time to weigh up possible courses of action to comply with logic standards.
Under such conditions, judgement made on basis of quick mental simulation of possible event outcomes (pros and cons of situation).
Person is ‘satisficing’ due to limited time and cognitive capacity.
Same heuristic process used when task knowledge limited.

Question 7

Recognition heuristic

Answer 7

Basis for judgements and decisions when knowledge limited (G et al, 99).
Study (Ayton & Onkal, 97): ppts groups according on knowledge of English football teams.
- people recognised cues in environment- didn’t need knowledge in order to make decent decisions- recognised major english cities, which also tend to have more successful football teams.
JDM based on recognition of relationship between cues in environment- things that may predict other things.
Recognition heuristic is ecologically-rational: based on familiarity of things that predict other things (not a ‘quick and dirty’ strategy used due to ignorance).

Question 8

Gigerenzer’s other ‘ecological’ heuristics

Answer 8

'Take-the-best' = using best cue from environment (eg. unemployment and homelessness rates). 
'Elimination-by-aspects' = choosing by discriminating among attributes (eg. deciding which politician to vote for based on candidates' stance on issues of interest).

Question 9

Revised approach

Answer 9

Gilovich, Griffin & Kahneman’s (2002) book:
Presented revised and strengthened approach.
Original approach generally too descriptive.
Cognitive mechanisms underlying heuristics need to be identified better through research.
But, original approach spawned lots of work in psychology and beyond, so was good.
No longer just ‘cognitive miser’ approach- heuristics more than just strategies to save cognitive effort when faced with uncertainty.
Link between heuristics (and biases) and dual-process accounts of rational reasoning- reasoning deemed as being (normatively) non-rational often useful successively in everyday life.

Question 10

2 types of Kahneman & Frederick (2002)

Answer 10

In revised account: come up with suggestions of two types of heuristics needed to explain bias and judgements.
‘Automatic’ and ‘deliberate’ heuristics (involve different cognitive processes).
- automatic- system 1 reasoning.
- deliberate- system 2.
Most ‘adaptive’ heuristics (eg. take the best) involve system 2 reasoning- strategies deliberately used to reduce cognitive load.
Original approach’s heuristics (eg. availability) involve system 1- automatic not used to reduce cog load.

2 automated heuristics:

1) Choosing by liking: deciding on basis of immediate, affective evaluation of options.
2) Choosing by default: deciding on basis of first option that comes to mind.

Question 11

Revised approach (cont).

Answer 11

Gilovich et al (02): more thorough treatment of issue of bias in cognition:

• shortcomings of original approach noted.
• Gigerenzer et al’s approach might only apply to system 2 thinking.
• but ‘recognition’ heuristic bears hallmarks of system 1 thinking (judging by familiarity), so more work needs to be done.

2 types of heuristic needed to explain bias in dual-process thinking (system 1 and 2)?
Interplay between cognition and feelings explains bias in certain situations- feeling states can guide judgements and decisions under conditions of uncertainty.

Using system 2 thinking to overcome bias:
- consider base-rates and conjunction probabilities.
- frequencies easier to understand than probabilities but format influences system 2’s operation.
Heuristics typically due to system 1thinking, but resulting bias moderated by system 2.

Question 12

Summary

Answer 12

Gigerenzer et al’s (99) ‘adaptive critique’:
- heuristics part of rational thought.
- ecologically-rational thinking that serves our needs to function in our environment.
- underpinned by ‘bounded rationality’ theory and notion of ‘satisficing’ (Simon, 1957)- rational thinking given cognitive limitations.
- normative models focused too much on probabilistic thinking.
Fast and frugal heuristics (eg. recognition) good for dealing with uncertain situations
Gilovich et al, 02 revise approach:
- originally approach too descriptive.
- link between heuristics (and biases) and dual-process accounts of rational reasoning- 2 types of heuristic needed to explain bias.
- interplay between cognition and feelings.