2.1 Cognition: Thinking & Remembering
Problem 7 – Estimating chances
Literature: Matlin, Robinson-Riegler
Heuristics
= general strategies; usually lead to correct solution
- Type 1: fast, simple, easily accessible heuristics; prone to errors
- Type 2: slow, correct original error, good decision
Representativeness Heuristic
= selected sample is similar to general sample population
- Representative = similar in important characteristics to original population
- e.g. THHTHT (1) seems more representative than TTTHHH (2)
o (1) seems more random; random like infinite amounts of coin tosses
o Random looking outcome seem more likely than orderly outcomes (21.97€ vs. 22.22 €)
→ sometimes random processes → outcomes that don’t look random
- Problem:
o Convincing, persuasive
o People ignore statistical info
Sample size
- Large sample → more likely to reflect true proportions of population
- Small sample → extreme proportion
(e.g. small hospital vs. large hospital; which is more likely to have 60% boy born)
- Small-sample fallacy = assumption that small sample is representative
o Stereotypes
Base rate
= how often something occurs in general population
- Base-rate fallacy = paying too little attention to important info about base rate
- E.g. description of student Max. Judge how likely which profession is. → judgements based on
description, not on number of people in profession
- Training → awareness and consideration
Stereotypes
= estimations based on few descriptions placed on a bigger population
- Study: 100 people in a room; 70 lawyers, 30 engineers; 2 descriptions
o Changed 70:30 probability to 50:50 after reading descriptions
o Description matched engineer→ adjustment of estimate
o Even when no evidence, people hope to improve predictions
- Baye´s Theorem = estimates based on 2 types on information
o Base rates of event
o Likelihood ratio (assesses usefulness of new info = diagnositicity)
Beliefs
- Confirmation bias = search/pay attention to evidence matching own belief; ignore rest
Salience and Vividness
- More salient and vivid → more attention, stronger impression
- E.g. flight crashes vs. car accidents → driving is more dangerous than flying, but doesn’t receive as
much media coverage
Conjunction Fallacy
= judge conjunction of two events as more probably than either one
- Conjunction rule = conjunction of two events can´t be more likely than each event separately
- E.g. bank teller (A), feminist (B)
Availability Heuristic
Problem 7 – Estimating chances
Literature: Matlin, Robinson-Riegler
Heuristics
= general strategies; usually lead to correct solution
- Type 1: fast, simple, easily accessible heuristics; prone to errors
- Type 2: slow, correct original error, good decision
Representativeness Heuristic
= selected sample is similar to general sample population
- Representative = similar in important characteristics to original population
- e.g. THHTHT (1) seems more representative than TTTHHH (2)
o (1) seems more random; random like infinite amounts of coin tosses
o Random looking outcome seem more likely than orderly outcomes (21.97€ vs. 22.22 €)
→ sometimes random processes → outcomes that don’t look random
- Problem:
o Convincing, persuasive
o People ignore statistical info
Sample size
- Large sample → more likely to reflect true proportions of population
- Small sample → extreme proportion
(e.g. small hospital vs. large hospital; which is more likely to have 60% boy born)
- Small-sample fallacy = assumption that small sample is representative
o Stereotypes
Base rate
= how often something occurs in general population
- Base-rate fallacy = paying too little attention to important info about base rate
- E.g. description of student Max. Judge how likely which profession is. → judgements based on
description, not on number of people in profession
- Training → awareness and consideration
Stereotypes
= estimations based on few descriptions placed on a bigger population
- Study: 100 people in a room; 70 lawyers, 30 engineers; 2 descriptions
o Changed 70:30 probability to 50:50 after reading descriptions
o Description matched engineer→ adjustment of estimate
o Even when no evidence, people hope to improve predictions
- Baye´s Theorem = estimates based on 2 types on information
o Base rates of event
o Likelihood ratio (assesses usefulness of new info = diagnositicity)
Beliefs
- Confirmation bias = search/pay attention to evidence matching own belief; ignore rest
Salience and Vividness
- More salient and vivid → more attention, stronger impression
- E.g. flight crashes vs. car accidents → driving is more dangerous than flying, but doesn’t receive as
much media coverage
Conjunction Fallacy
= judge conjunction of two events as more probably than either one
- Conjunction rule = conjunction of two events can´t be more likely than each event separately
- E.g. bank teller (A), feminist (B)
Availability Heuristic
