Confirmation Bias

seeking confirming information
- Lindholm & Christianson (1998) (n = 92)
- rather blame innocent immigrant for crime than innocent own folk
  - only within own folk, not within immigrants
how to circumvent:
- switch perspective: what if x is the other way round, can be astonishing

Influence of Schemas

bank robbery: schema relevant information is seen, the rest not
- Tuckey & Brewer (2003) (n = 106)
- color of gun is seen, color of getaway car is not
- replication study: even with a ski mask the gender of the robber is mostly male
who is smart?
- Darley & Gross 1983 (n = 70)
- social class has impact on memories of individual
- seeing Hannah in the poor environment changes the memory of her test score negatively

asking a different question changes the outcome (Loftus & Palmer 1974)
- cars have smashed/collided/bumped/hit/contacted
- collided/crashed cars were faster
telling people what happened makes them confirm a memory
- even negative memories
  - 70% can report on false memory of having police contact in adolesence (Shaw & Porter 2015)
misinformation can “create” memories which never happened
- making up that one has visited Disneyland in childhood (Braun et al 2002)

extent to which knowledge is activated in memory, thus likely to be used in cognition and behavior
e.g. weapon focus effect

Medvec, Madey & Gilovich (1995)
- 3rd on podium is happy … made it onto the podium
- 2nd on podium is unhappy … did not get 1st place
overconfidence
- overestimating accuracy of memories or predictors
- eyewitness testimonies
  - small correlation of accuracy and confidence in memory
  - just because eyewitness is more “sure” it is not more likely to be correct
Overestimation vs Overplacement vs Overprecision
Moore & Healy (2008)
- n = 82
- overconfidence in simple tasks, too little confidence in complex/difficult tasks
  - reminds he of “you overestimate what you can do in a year, but underestimate what you can do in 10 years”
- can you program? → not confident enough, thinking they are worse than others
- can you use a computer mouse? → too confident, thinking they are better than others
overconfidence & exams
- Miller & Geraci (2011)
- n = 113
- top half of students have better predictions about their performance
  - but also higher confidence
- Quasi Experiment bc we cannot change the confidence of the students