Biases Flashcards
Availability Bias
The easier (or faster) we remember an example of an event, the more likely we believe it is
Example Availability Bias (9/11)
Right after 9/11, people were flying less and driving about 3 % more. This extra driving was responsible for 353 excess deaths in the three months following 9/11
Hindsight Bias
When some event occurred, people think in retrospect that they have evaluated the event to be more likely (ex-ante) than they really did
Bias is particulary strong if small probabilitites are involved
Insufficient sample size
Humans tend to base their judgements on excessively small samples, often even on individual cases
E.g. „Four out of 5 dentist suggest that … —> does not indicate how many dentists have actually been surveyed
Representativeness Heuristic
The probability that some object belongs to some category is evaluated to be higher if the object looks representative for the category
representativeness is an assessment of the degree of correspondence between a sample and a population)
E.g. Criminal justice: Jurorsmay make judgments about guilt based on how closely a defendant matches their prototype of a guilty suspect
–>Can contribute to systematic discrimination
Insensitivity to base rates
People tend to give too little weight to the base rate (i.e. prior probability in the context of Bayes Theorem
is a direct consequence of the representativeness heuristic
Reversing conditional probabilitites
People tend to confuse conditional probabilitites and minsinterpret the direction of the conditioning for a conditional probability
reversion conditonal probabilitites is common + has serious consequences
Conjunction Fallcy
A compounded (and thereby more specific) event is evaluated to be more likely than a single general one
E.g. Volleyball Players age and height
Anchoring and adjustment
Bias: Adjustment is often insufficient, and the anchor exerts too much influence, which causes the estimates to stay too close to the anchor
- Anchoring: A “near at hand” first guess serves as an anchor point
- Further thinking causes adjustment of the original guess
Overconfidence Bias
The tendency of a person to place too much confidence in his own abilities , knowledge, or the quality of his forecasts
Overview Systematic errors and biases
Incomplete or inappr
Incomplete or inappropriate data:
- Availability bias
- Hindseight bias
- Insufficient sample size
Incorrect adjustment/processing of probabilitties:
- Representativeness Heuristic
- Insensitivty to base rate
- Reversing conditional probability
- Conjunction fallacy
Insufficient critique of one own judgement:
- Anchoring and adjustment
- Overconfidence
What is the Problem of Uniform prior interpretation?
Each elementary event has the same chance of occuring. With n events 1/n
Problem: For decision situations beyond the ralm of gambling, this kind of probability interpretation does not yield anyhting at all.
What is frequentist interpretation?
Frequentist interpretation: The probability of future occurences of an event is deducted from the relative frequency
- Prob that smokers will contact lung cancer is deducted from the relative frequencies observed in the past
What is the problem with frequentist interpretation?
In many situation in which difficult decisions have to be made, it is not possible to rely on relative frequencies because the situations are new
