Group Decision-Making

Question 1

group decisions

Answer 1

can be very consequential

Question 2

downsides of group decision-making

Answer 2

• conformity
• authority
• polarisation
• groupthink

Question 3

benefits of group decision-making

Answer 3

• wisdom of the crowd
• improved reasoning

Question 4

conformity

Answer 4

how people’s judgments can be influenced by those around them, social pressure to conform (fit in)

Question 5

conformity study

Answer 5

Asch 1956
- line length experiment
- confederates (6/7) gave the wrong answer on purpose even though correct answer was obvious
- 75% of real p’s then gave wrong answer on at least one trial, as the other ‘participants’ did too
- social pressure of being the only person to give a different answer caused p’s to conform to the majority view (may have knew the correct answer despite this)

Question 6

authority

Answer 6

authoritative figures can be important in order for groups to function, but can also lead to poor communication between authoritative figures/ ranks communicating to each other

Question 7

authority in aviation

Answer 7

• Tenerife 1977 plane crash
• KLM and Pan-Am jets crashed into each other on the runway
• due to engineer not challenging the captain’s decisions to take-off (due to captains experience), despite doubting that the runway was clear
• unwillingness to challenge authority lead to many deaths

Question 8

authority report

Answer 8

Alkov et al 1992
- found that the bigger the difference in rank/seniority between crew members in a cockpit, more likely an incident was
- 40% of junior co-pilot reported not relaying concerns about safety to senior pilots

• gave rise to crew resource management to train people to challenge seniors, and for seniors to recognize this

Question 9

polarization: the risky shift

Answer 9

• polarization came from the ‘risky shift’: Stoner et al. 1961.
• found that groups endorsed a riskier judgment than the average of individuals by themselves (the risky shift)
• opposite can also occur with more cautious approach to be endorsed in a group

Question 10

polarization

Answer 10

groups will tend to endorse a more extreme view than the average of its members’ views

Question 11

polarization in real life

Answer 11

• juries
• Myers & Kaplan 1975
• p’s considered 8 felony cases (4 very incriminating, 4 not very incriminating)
• p’s judged defendant’s guilt on all cases on a scale as individuals, then in a group
• p’s also asked to state punishment they would endorse on a scale individually, then in a group
• showed polarization: rating of guilt or innocence in groups became more extreme, same for severity of punishment ratings
• maybe being in a group and having people agree with them makes them become more confident in judgment and endorse more extreme view

Question 12

groupthink

Answer 12

Irving Janis 1981
- when groups are too homogeneous, members become un-receptive to critique and even actively defend the consensus (general agreement) from outside critique

Question 13

examples of groupthink going wrong

Answer 13

• Bay of Pigs invasion
• The Challenger launch - NASA failed to take a recommendation from engineers that the temperature was too low for a safe launch: were not receptive to criticize

Question 14

group think characteristics

Answer 14

• cohesive groups striving for unanimity and avoiding conflict or criticism
• exacerbated in homogeneous teams whee members are similar

Question 15

group think downsides

Answer 15

• too loosely defined
• when decision making goes wrong we may call it this, but probably more down to conformity, authority and polarization co-occurring

Question 16

wisdom of the crowd

Answer 16

• Galton 1907
• the average estimate of a group of people is highly accurate, often better than the individual estimates
• calculated average estimates of people who entered a competition to guess an ox’s weight

Question 17

research on the wisdom of the crowd

Answer 17

Kurver et al. 2016
- examined how doctors performed when classifying mammograms when cancerous or not
- average accuracy was better than an individual doctor’s accuracy
- BUT depends on variability of doctor’s accuracy
- doctors with similar levels of diagnostics accuracy will outperform the best doctor in the group with the average
- if there is low level of similarity between doctors level of diagnostic accuracy, they will not do better than an individual doctor (the best in the group)

Question 18

a judgment where wisdom of the crowd can fail

Answer 18

surprisingly popular

Question 19

surprisingly popular

Answer 19

Prelec et al.
- done to find the correct answer/ niche or specialized knowledge
- ask what their answer is and what they think most people will answer?
- if you know your knowledge is niche and predict that people won’t agree with you, then that is the answer you should choose
- correct answer will have a higher actual popularity (what people chose) than expected popularity (what people thought others would chose)

Question 20

wisdom of the crowd (within)

Answer 20

Herzog and Hertwig 2009

• can one individual come up with estimates, average the estimates to improve individual performance?
• p’s estimate dates of historical events
• 2 conditions:
  1) Reliability condition (simply asked to provide another estimate without being asked to alter reasoning)
  2) Dialectical bootstrapping condition (arguing with themselves and coming up with new estimates based on alternative reasoning)
• using alternative reasoning to argue for different estimates lead to 4% improvement in performance once estimates were averaged (wisdom of the crowd in one person)

Question 21

wisdom of the crowd within critique

Answer 21

averaging estimates with just one other person outperforms dialectical bootstrapping (for an individual) - better to team up with other people

Question 22

summary of wisdom of the crowds

Answer 22

• effective at cancelling out random errors (produced by some people)
• not so effective at cancelling out systematic errors (e.g. anchoring) (so bias in estimates)(but potentially overcome in some cases)

Question 23

Sasaki & Pratt

Answer 23

• attraction effect experiment
• study with ants choosing between potential nest sites
• individuals are biased by the presence of a decoy option, ants chose the nest that dominated the decoy
• however with colonies of ants there was no decoy effect, suggesting that collective decision making can overcome systematic errors in bias (collective rationality)

Question 24

improved reasoning from group selection task evidence

Answer 24

Moshman & Geil
Wason card selection task
- individuals chose P and Q card (9% chose P and not-Q_
- groups more likely to pick P and not-Q cards (75%)
- in groups none of the members initially selected the correct cards, but with group reasoning it improved performance

Question 25

improved group reasoning more evidence study - numbers-to-letters reasoning task

Answer 25

Laughlin P's must decipher which letters A to J map onto numbers 0-9 Performance measured by number of trials taken to identify all mappings Number of letters in proposed equations also measures performance - groups used more letters in equations - less trials taken on group task - improved performance in groups of more than two people (not pairs), statistically significant

Question 26

common themes about good and bad things for group reasoning

Answer 26

independence and diversity - necessary for success in group decision making