Nick Chater Behavioural Science Group OVERVIEW THE FICTIONAL SELF THE ILLUSION OF DEPTH WHY BEHAVIOUR IS SOMEWHAT COHERENT IMPLICATIONS 1 THE FICTIONAL SELF The claim We infer our own inner life from our words and actions just as we infer those of a third person ID: 782485
Download The PPT/PDF document "The Fictional Self: Implications for The..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
The Fictional Self: Implications for Theories of Decision Making
Nick ChaterBehavioural Science Group
Slide2OVERVIEW
THE FICTIONAL SELF
THE ILLUSION OF DEPTH
WHY BEHAVIOUR IS (SOMEWHAT) COHERENT
IMPLICATIONS
Slide31.
THE FICTIONAL SELF
Slide4The claim
We infer our own inner life from our words and actions, just as we infer those of a third personA popular viewpoint in many areas of social psychology and philosophy Corollary 1:
We
do not
have any direct access to “inner” beliefs, desires etc...
Corollary 2:
Indeed, there
are
no such beliefs and desires (illusion of depth)
Slide5Inferring our own beliefs 1
Festinger and Carlsmith, 1963
1. People do a boring task
2. Then paid $1 or $20 to persuade others to do it
People
paid more
find it
more aversive
; less likely to do it again, etc.
Why did I do this?
“If I was only paid $1, it can’t have been too bad”
(Daryl
Bem
self-perception interpretation, not quite
Festinger’s
)
Slide6Inferring our own beliefs 2
Reber & Schwarz 1999“fluency” of reading (!) affects plausibility of a statement
Osarno
is a town in Chile
Osarno
is a town in Chile
But effect is cancelled when people have an alternative explanation (poor photocopier)
Alter and Oppenheimer 2006
IPOs with ‘fluent’ names do substantially better (though effect washes out) RDO
vs
KAR
Slide7Inferring our own beliefs 3
(Schwarz et al)Generate 3 reasons why...
Happy with decision,
like partner
3 is
easy
;
high
‘fluency’
infer
strong
commitment
high self-report
on happiness with decision, partner, etc.
Generate 6 reasons why...
Happy with decision,
like partner
6 is
hard
;
low
‘fluency’
infer
weak
commitment
lower self-report
on happiness with decision, partner, etc.
Slide8Inferring our own emotions 1
Schachter and Singer (1962) Epinephrine or placebo injections
Raising (or not) arousal
informed (or not) about side-effects
Put with euphoric or angry ‘stooges’
Higher arousal
participants are more euphoric/angry
When people can ‘explain away’ arousal, effect much reduced
Slide9Inferring our own emotions 2
Dutton and Aron (1974)
Male/female interviewers of
male bridge-crossers
low
vs
high bridge
Interviewers give phone number
“further explanation” of study
More calls to
female
interviewers
for
high
bridge
Arousal (caused by high bridge) misattributed as caused by attractiveness
Slide10Inferring our own preferences 1
Johansson, Hall et al., ScienceFalse feedback on choices
not noticed
rationalization given
later preferences changed
And it works with jam
Slide11Inferring our own preferences 2
Can people stably determine preferences between pains and money?Or do they have little idea what is a “reasonable” trade-off?BDM auction with small electric shocks
Vlaev
, Seymour, Dolan &
Chater
, Psych Science, 2009
Slide12Slide13Pain magnitudes were presented in pairs in three blocks of ten trials
Two “endowment” conditions
£0.40 per trial
£0.80 per trial
Slide14Slide15People double their offers, when they have double the money...
Value of pain changes by x2 within minutes!
Slide16In a slogan:
We are fictional characters of our own creation
i.e., the body isn’t fictional; but the ‘character’ is!
Daily life (and decision making) is a bit like improvisational acting
“What would X do now?”
Slide172.
THE ILLUSION OF DEPTH
Slide18Where does a rocket take off?
Triangulation will solve it; and the more measurements, the more error can be reduced
Slide19Where does a rainbow touch down?
Triangulation won’t work...
And the problem is
not
eliminating noise!
Slide20Rainbows, like beliefs and values have the
illusion of depthBut different ‘bearings’ don’t converge
Because they are improvised on the fly
Not products of stable, (if somewhat hidden, motives)
Different methods of elicitation will yield
irrenconcilable
“bearings”
Slide21Reason-based choice (Shafir
et al)Expensive, but exciting holiday
vs.
Cheap, but dull holiday
Bali
Bournmouth
Slide22Which holiday would you
choose?
Bali
Bournmouth
Reason: “It’s exciting!”
Slide23Which holiday would you
reject?
Bali
Bournmouth
Reason: “It’s too expensive!”
Slide243.
WHY BEHAVIOUR IS (SOMEWHAT) COHERENT
Slide25But behaviour is by no means wholly incoherent
Rationality constraintsRules (moral or other)
Herding and self-herding
(what do I/we normally do?)
Hierarchies of goals
Prices (multiplicative
additive)
What does my character do next?
Slide26Rationality constraints
£1 + £1 + £1 + £1 is not really preferred to £20 - £15 (despite prospect theory)
We don’t always break a diet when passing the fridge (despite temporal discounting)
We don’t actually fall for Dutch books
Rational choice theory helps capture
qualitative patterns of reasoning
that govern the ‘stories’ we create about ourselves (or others)
Rationality constraints on “story generation” are a (
partial
and
local
) stabilizing force
Slide27Prices
Decision by arguments (cf Loomes):
Comfortable
vs
uncomfortable seating
Cost £X
vs
£Y
strength of ‘argument’ driven by ratio X:Y
Duration tends to be forgotten about
How much would you pay for 1 h of sitting (more) comfortably?
Or 4h 50m of sitting (more) comfortably?
Slide28EVERYTHING IS RELATIVE...
2
nd
class: £164.10
Business: £325.10
Second: £108.30
First: £229.90
How much would you pay for 1 h of sitting (more) comfortably?
Or 4h 50m of sitting (more) comfortably?
London to Edinburgh, return
In each case,
double
the basic price: comparison in action
Slide29By being able to ‘unbundle’ enforces additivity
(and thus more stability)Bundled -> multiplicativeHouse size and location
Car quality + stereo system
Doubling
of all wine prices in a restaurant
Taking a
percentage
of sale in auctions, house sales, mergers
Same good: but will happily pay wildly different amounts
Unbundled
Computer hardware + software
Unlicensed restaurant; bring your own drinks
The supermarket, rather than the restaurant.
Same good, same price (but not via same
utility
)
Slide304.
IMPLICATIONS
Slide31IMPLICATION 1: THE POWER OF THE STATUS QUO
Current house prices soon seem ‘normal’
Wage increases rapidly ‘fade out’ as contributors to happiness; And even lottery wins
Tax and other cost changes have
much
less long-term impact on behaviour than is typically expected
Levels of risk are viewed relatively
: sub-prime lending, complex instruments etc judged
against each other
Slide32IMPLICATION 2: THE POWER OF THE HERD
No absolute anchors
Comparison against other’s beliefs and
behaviors
Risk-takers will be
relatively
risk seeking,
w.r.t
., peer group
Slide33IMPLICATION 3
UNANCHORED PRICES
“value”
“price”
PEOPLE KNOW THE PRICE OF EVERYTHING, BUT THE VALUE OF NOTHING
People can’t “know” their values
So they must partly
infer
them from market prices
Allowing feedback loops between values and prices
The origin of booms and crashes?