Tversky-Kahneman
and Co.
availability
overconfidence
anchoring
representativness
"law of small numbers"
people want sample to fully represent properties of distribution
and see true randomness as trends and order
ignoring Bayesian prior probabilities
same variance for human generated sample of 10, 100 and 1000
prospect theory
herding
"peer effects" and Granovetter models for social action
priming
adjustment to anchor
unexpected unavailability (negative second derivative)
the more supporting evidence you have to create, the less confident you become
easy to recall = frequent
personal = even stronger
true randomness looks like tendency to cluster
can be by completely irrelevant factors, meaning we all do that all the time :)
risk
affect heuristic
substitute “what do I think” with “how do I feel”
associative coherence
we don’t like results that are against our feelings/views
that might be all you need to explain home market bias
paul slovic
conjunction falacy - subset more likely than superset only because it matches the description more closely (coherence over logic)
less is more
joint valuation actually compares values, where single valuation regresses to average
10 high value baseball cards valued lower than same 10 + 3 low value in single sided valuation
this links perfectly to non-linear valuation like UVM
causal over statistical
causal base rate has more weight with people than statistical, even if they actually say the same
helping experiment - dissemination of responsibility
learning psychology
"subjects' unwillingness to deduce the particular from general was matched only by their willingness to infer general from the particular"
surprising individual case matters more in changing opinion than actual statistic
regression to the mean
causal interpretation of random effects
substitution - easy and similar question in place of a true hard one
intensity matching
prediction = evaluation of current evidence, not accounting for uncertainty or reliability of evidence
coherent causal stories
hindsight bias
outcome bias
planning fallacy
inside view vs outside (baseline)
too optimistic forecasts/budgets/decisions
failure to see risks, overfocus on own planskills
loss aversion
decision weight vs probability (steeper at the extremes)
overestimating low probability (mostly based on cognitive ease and availability) vs underweighting based on lack of availability
denominator neglect - 1 in 1000 will be taken as higher/more important than 0.1%
sunk-cost error
fail often,fail fast
Thaler
transaction utility - bargain vs rip-off
reference price (contextual)
acquisition utility
purely economic is utility of having A above price
sunk cost fallacy
mental accounting
mental accounts are topical
house money effect
break-even risk taking in losses
self-control
present-focused time-inconsistent over exponential discounting which is time consistent
Colin Camerer - behavioral game theory