Judgements and probabilities - RD3 Flashcards
What is a judgement?
calculating the likelihood of events using incomplete information (e.g 68% likely to happen)
How are judgements different from decisions?
judgements are the probabilities of each option being correct, decisions are the actual option you choose (usually the one with the highest probability)
what are hits and correct rejections also known as?
hits = sensitivity
correct rejection = specificity
What do people aim to do when creating tests for diseases?
maximise hits and correct rejections
Even if a test is the same accuracy, what can affect the chance that someone with a positive result actually has a disease?
base rate
What is Bayes rule? what do the components mean?
Posterior ∝ likelihood x prior
- posterior = updated belief that you have a disease (given the data)
- likelihood = probability of the test result give prior
- prior = base rate (probability of having the disease)
what is the fraction used to calculate posterior odds?
the probability of not having the disease given the evidence/the probability of having the disease given the evidence
what is the fraction used to calculate likelihood ratio?
the probability of getting a positive result given that you don’t have the disease/the probability of getting a positive result given that you have the disease
what is the fraction used to calculate the prior odds?
the probability of not having the disease/the probability of having the disease
What numbers from the efficacy of a test do you use when working out the likelihood ratio?
the false alarm rate and the hit rate
what numbers do you use when working out the prior odds?
The prevalence of the disease and 100- the prevalence
How does the engineer/lawyer study by Kahneman and Tversky (1973) show base rate neglect?
Participants thought the person was more likely to be an engineer after hearing about his character, while ignoring that most people in his group are lawyers so logically he’s more likely to be a lawyer
What did cascells et al (1978) find when they told medical students the prevalence of a disease and the false positive rate?
- base rate neglect
- just used the false positive rate to work out how likely they would be to have the disease with a positive result
What is the conjunction fallacy?
the mistaken assumption that the probability of the conjunction of 2 events is greater than the probability of one of them
Why might we fall for the conjunction fallacy?
we value precision/specificity and assume that information in both options is redundant
What did Tversky and Kahneman (1983) find in the bank teller problem?
people said that linda was more likely to be a bank teller who is in the feminist movement than just a back teller
conjunction fallacy
Even though many say we aren’t optimal, what things can improve our performance? (3)
- including info about causal structure
- presenting in frequentist terms
- personal relevance to the problem
What is the representativeness heuristic?
assume an object/individual belongs to a specific category because it is representative
What is the availability heuristic?
frequencies of events are estimated by ease of retrieval
What did Lichenstein (1978) find when looking at how commonly people rated causes of death?
Publicised deaths were rated as more common - e.g murder and suicide
What are the biggest drivers for how commonly people will rate a disease according to Pachur et al (2012)? (3)
- direct experiences (biggest driver)
- emotional response (affect heuristic)
- media coverage
What are 4 problems with heuristics as a theory for how we judge?
- they are vaguely defined
- doesn’t define when specific heuristics are used
- not necessarily biased processing but poor information
- a list doesn’t equate to a ‘theory’
What is the dual-process theory?
judgements are based on 2 distinct systems: system 1 (fast, automatic, effortless, implicit) and system 2 (slow, serial, effortful, controlled)
In dual-process theory, when do we use the different systems? What is the problem with us and our use of the systems?
use system 1 most (heuristics) - but this is not optimal and is prone to errors
cognitive misers - we can use system 2 to get the right answer but will often use 1 because it’s easier
What are some problems with the dual-process theory? (4)
- system 2 also leads to errors sometimes
- could be more of a continuum, not a hard distinction
- both systems are relatively ill-defined
- some evidence for intuitive use of base rate (system 1)
Which rules do we use when deciding which city is bigger? (3)
- search rule - do I know the city or anything about it?
- stopping rule - stop search if you only know one city
- decision rule - choose the city you know
How did telling people a city had an airport change their estimates of its size compared to a known city? What does this tell us?
Mostly still based on recognition but not as much as if they weren’t told about the airport
so they used relevant information to help them
What is the correlation between the validity of the information of knowing a city and its usage?
.64
don’t go on recognition as much if that’s not helpful, like judging the northness of a city
