W7 Probabilities ✅

Question 1

What are the factor that influence wrong probability judgements?

Answer 1

Heuristics: shortcuts that reduce effort but are prone to bias/error
-> 3 types of heuristics strategies: availability, representativeness and anchoring

Heuristics strategies are based on ecological rationality: Apparent biases from rational responses using ecology (context-clue) Ps operate in
-> 2 phenomena: natural frequencies, misperception of randomness

Question 2

What are supporting evidence for availability heuristic?

Answer 2

1. Estimates causes of death: people overestimate rare events (e.g. deaths by sharks) and underestimate common events) -> more headlines for deaths by sharks
2. Effect of memory: names of famous group are easier to remember, especially for gender with more famous names
3. Conjunction fallacy: two events happened at the same time will be more likely to happen together than each event happen separately.
   -> words with ‘—ing’ is more frequent than words with ‘—-n-‘

Question 3

What are supporting evidence for representativeness heuristic?

Answer 3

1. Judgments of probability: based on assessments of similarity to existing figures
2. Base rate neglect: based judgement on the description, ignoring base rate (population probability)

Question 4

What are supporting evidence for anchoring heuristic?

Answer 4

Statistics are taken at random (Ps spin the wheel) AND rigged wheel to give either 10% or 65%
=> Ps’ guesses to a question based on the random number they got (irrational)

Explanation: anchor number adjusted the estimate people made as opposed to figuring out true value

Question 5

What are 2 biases due to ecological rationality and their explanations?

Answer 5

1. Natural frequencies
2. Misperception of randomness

Question 6

What are supporting evidence for the misperception of randomness bias?

Answer 6

• Gambler’s Fallacy: people expect a long sequence to be randomly generated.
  -> e.g. HHHHHT is not as representative as HTHTTH (coin flip)
  -> even when probability is still 50/50 whether the streak continues or not.
• Past experience: Inappropriate generalization of past experience
  -> Random mechanical outcomes = Sampling without replacement
• Memory constraints assumptions:
  -> People only ever see finite sequences
  -> People can only hold a short subsection of a sequence in memory (e.g., the last 4 outcomes)

Question 7

What is the supporting evidence for natural frequencies bias?

Answer 7

1. Natural frequencies: people are better at tracking natural frequencies (numbers > percentage of probability)
   -> Example: guess probability of breast cancer with positive mammogram
   => Bayes’ equation allows us to update estimated probability based on existing and new knowledge.