Decision-making: Tuesday July 16, 2013
Cognitive miser: we are cheap with the way we distribute mental resources to decision
making due to time pressures.
What is decision making?
Sources of decision-making difficulty:
1. Conflict = decision makers must make tradeoffs
2. Uncertainty= outcome of decision often depends on uncertain variables or
Phases of decision marking:
1. Setting goals
2. Gathering info
3. Structuring the decision – weigh different options
4. Making a final choice
5. Evaluation – reflect upon the decision making process, what was critical in
the decision? How can I improve my choices in the future?
Heuristics in decision-making: mental shortcuts we use when making decisions
that can lead to biases. People: Daniel Kahneman and Amos Tversky.
Availability heuristic: in 4 pages of a novel (2000 words) how many words would
you expect to find having the form (one or the other):
----ing most people would chose this one.
-----n- but this could include ing words…so its this one.
The words that end in ing are the ones more available in memory, therefore
they chose that one. The automatic (implicit) cognitive system makes
decisions that are wrong.
Adding on addition clauses makes something statistically less plausible.
Ex. Ross and Sicoly: 1979
Ask people to estimate the extent of household chores each person does
(since you think back to the last thing YOU did, you’ll always think it was
you) ego centric
Both think they did more
Well publicized risks tend to be overestimated. - RISK PERCEPTION
**Weigh things more in decision-making that your memory attends to more.
Might be attended to more
Of all families having exactly 6 children, in what % do you think the exact birth
order of boys or girls was? a. ) BBBGGG Although either one could be chosen because with each
pregnancy there is a 50/50 chance. So since this one has a seeming pattern to
it, they chose it less often.
b. ) GBBGBG chose this one, because people want to see randomness.
**Random things should look random, it’s the expectancy we push on something.
Might be more stereotypical
Law of small numbers: People have mistaken belief that small samples are
representative of the population from which they are drawn.
8 vs. 100,000 coin flips
Which is more likely to reflect the actual odds? (50/50) Distribution is not affected
Gambler’s fallacy: You are more likely to see a red after which of the following
a. BBBBBBB? People chose this one, but it doesn’t matter, what happened in
the past should not even be considered, like with pregnancy example above,
each time there is an equal 50/50 chance.
Group 1: 8x7x6x5x4x3x2x1
Group 2: 1x2x3x4x5x6x7x8
Ask for ballpark number. Findings: Despite the fact that the question is the same,
group one had mean estimate of 2250, and group 2had 512. But the actual answer is
40,320. When people see the bigger number early on, they prime for bigger ones,
and anchor to a bigger answer, and vice versa for the 2 group.
Example – consumer choice: 4.99 vs 5.00 – they are trying to anchor people to the
lower number than the higher one.
Framing: Glass half full or half empty.
5-cent charge for credit vs. 5 cent discount for cash: Its exactly the same
consequence but we’d like one more than the other.
600 people affected by a deadly disease:
Positive treatment a: save 200 treatment b: a 33% change of saving all, or 66% to
kill everyone 72% CHOOSE A
Negative treatment a: 400 will die treatment b: a 33% that no one will die, 66%
will all die 78% CHOOSE B
How you frame it makes the difference, because they are exactly the same.
Loss aversion: people prefer avoiding losses to making gains
Endowment effect: value something more when you have it in your possession, mug
was overvalued when you’ve been endowed with it
***Illusory Correlation: seeing a relationship you expect to see even when none
exists. Confirmation bias: 2-4-6
-Given 3 cards that adhere to a pre-determined rule and you’re asked to tell them
the rule. You can make new examples and ask to see if they satisfy the rule.
The correct rule is that generally the numbers are increasing.
- Findings: 29 original participants, 6 found it without any other tries, 13 made one
wrong guess, 9 reached 2 or more wrong ones, and 1 made no conclusion at all
- People develop an idea of the rule then make examples that follow the rule they
FAIL to make a counter example (one that even though it follows their rule it won’t
get a yes from the researcher) See if they can falsify it.
- They try to confirm their rule, instead of testing the rule.
Selection task (confirmation bias cont’d:
- Clerk if an a on one side, a 4 should be on the other