Judgement under Risk + Uncertainty Flashcards
What is a Heuristic?
Rule of Thumb / Mental Shortcut
- Reduces Effort to make Judgements – BUT can lead to Bias
What are the 3 main Heuristics?
Representativeness
Availability
Anchoring + Adjustment
What are the 2 components of the Representativeness Heuristic?
Insensitivity to Prior Probability
Insensitivity to Sample Size
What are the Rules of Probability?
Addition (Disjunction) Rule
Multiplication (Conjunction) Rule
What is the Disjunction Rule?
Addition Rule
Pr(A or B) = Pr(A u B) = P(A) + P(B) - Pr(A n B)
How do you show that A and B are Independent?
Pr(B|A) = Pr(B)
How do you show that A and B are Mutually Exclusive?
Pr(A n B) = 0
What is Pr(not-A)?
1 - Pr(A)
What is Bayes Rule?
Pr(A|B) = [Pr(B|A) x Pr(A)] / Pr(B)
=> [Pr(B|A) x Pr(A)] / [Pr(B|A) x Pr(A) + Pr(B|not-A) x Pr(not-A)]
What is the Representativeness Heuristic?
Probabilities are evaluated to degree that A is Representative + Similar to B
What is Prior Probability?
(Base Rate)
Has no effect on Representativeness - BUT should have Major Effect on Prob.
- Considerations of Base Rate freq. do NOT affect similarity of a description to the stereotypes of Engineer or Lawyer
– If ppl Evaluate Prob. by Repres. –> Prior Prob. will be Neglected
What is Base Rate Neglect?
Failing to consider Base Rate –> Over/Underestimate
Representativeness –> Neglecting important parts of Bayes Rule - e.g. Base Rate - p(Hypothesis)
What is Insensitivity to Sample Size?
Ppl assess Likelihood of Sample Result by Similarity of result to the Population Parameter
- Fail to appreciate Role of Sample Size
Representativeness –> Judged Prob. of Sample Statistic will be Independent of Sample Size
What is the Law of Large Numbers?
Over large no. of Trials - Empirical Prob. of an event will approach its True Prob. (Exp. Value)
- Very large samples from a Prob. Dist. very accurately reflect Prob. in the Distribution
What is the Law of Small Numbers?
Sampling Theory –> Smaller Sample MORE likely to deviate from True Prob.
- Ppl mistakenly apply Law of Large no. to Small Samples
Representativeness –> Misconceptions of Importance of Sample Size
Overgeneralise from Small samples to Dist.
What is Gambler’s Fallacy?
Mistaken belief that 2 Outcomes are Dependent when they are NOT
What is the Consequence of LoSN in Gambler’s Fallacy?
False belief Sequence of Independent draws from a Dist.
- Outcome not occurred in a while so is more likely to come up next draw
What is Hot Hand Fallacy?
2 Independent events are Considered Dependent
- Player’s chance of hitting a shot is Greater following a hit than a miss on previous shot
What is Conjunction Fallacy?
Pr(A n B) is beloved to be Greater than P(A)
- Similarity between Linda + a Feminist Bank Teller > Similarity between Linda + Bank Teller
- Similarity Relations do NOT follow Conjunction Rule
What is Anchoring?
Ppl begin process of Estimation w/ whatever info readily appears in their minds
- Initial Anchor heavily influences final estimates
What is Adjustment?
Reassess the initial answers based on rough notions of what is a not-too-silly answer
What is System 1?
Intuitive System uses Heuristics that sometimes get things wrong from POV of Conscious Reasoning
- It is adaptive
What is System 2?
Conscious Reasoning slow but a Sophisticated process
- Very flexible + can be changed + improved by learning
- BUT can only concentrate on 1 thing at a time + requires effort to control
What are the unintended side effects of generally Adaptive Processes?
Errors + biases in Judgement
How do ppl tend to estimate Conjunctive + Disjunctive Events?
Overestimate Prob. of Conjunctive Events
- Overall Prob. < Success Prob. at each stage
Underestimate Prob. of Disjunctive Events
- Overall Prob. > Success Prob. at each stage
What is the Availability Heuristic?
Assess freq. of a Class/Prob. of event by ease w/ instances/occurrences that can be brought to mind
– Classes whose Instances are more easily retrieved –> Overestimates
