Thinking Fast and Slow - Study Notes - Version 2
Saturday, May 31, 2014
00b - Author's Introduction
Introduction
- purpose: improve vocab. for talking about judg’ts & choices of others, the company’s new policies, a colleague’s investment decisions. — ∃ distinctive patterns in the errors people make, systematic errors, biases
- the mental work that produces impressions, intuitions, & many decisions is unconsc.
- most of our judg’ts & actions are appropr. most of the time — but not always
- Aim: improve ability to identify, underst. errors of judg’t & choice — in others, eventually in selves, provide richer, more precise lang. to discuss them
Origins
- current underst. of judg’t & decision making ß psychological discoveries of recent decades
- collaboration w. the late Amos Tversky, many years
Where We Are Now
- to present a view of how the mind works that draws on recent devep’ts in cognitive & social psych’y —we now underst. intuitive thought, its marvels & flaws
- one kind of valid intuition ß experts have learned to recog. familiar elements in a new situation, act approp’ly — good intuitive judg’ts come to mind immed’ly — w. relevant expertise, recog. situation, intuitive sol’n likely correct — intuition may also work when difficult question & skilled sol’n available
- emotion very impt. in intuitive judg’ts & choices
- the essence of intuitive heuristics: faced w. a difficult question, we often answ. an easier question instead, usually without noticing
- spontaneous search for intuitive sol’n sometimes fails —no expert sol’n & no heuristic answ. comes to mind — then may switch to a slower, more deliberate, effortful form of thinking, i.e. “Slow Thinking
- “Fast Thinking” incl. both variants of intuitive thought, the expert & the heuristic, also auto. perception & memory
What Comes Next
·
Part 1. Two Systems : basic elements of two-systems approach to judg’t
& choice — elaborates distinction of auto. operations of System 1 vs.
controlled operations of System 2, shows how associative memory ( the
core of S1) continually constructs a coherent interpr’n of what is going on in
our world at any instant — the complexity & richness of the auto. & often
unconscious processes that underlie intuitive thinking, how they explain the
heuristics of judg’t — a goal is to introduce a language for thinking & talking
about the mind
·
Part 2. Heuristics and Biases: updates
the study of judg’t heuristics — asks: Why is it so difficult for us to think
statistically? — we easily think associatively,
metaphorically, causally — but statistics req. thinking about many things at
once — S1 is not designed for that
·
Part 3. Overconfidence: our excessive
confid. in what we believe we know, inability to acknowl. our ignorance &
uncertainty of the world we live in —illustr. by difficulties of statistical
thinking —prone to overestimate how much we underst. the world, underest. the
role of chance — overconfid. increased by certaintyof hindsight (illusory)
·
I hope for watercooler conversations
that intelligently explore the lessons that can be learned fr. the past while
resisting the lure of hindsight & the illusion of certainty
·
Part 4. Choices: in economics, the
nature of decision making, & assumpt’n
that econ. agents are rational —using the two-system model, the key concepts of
prospect theory (our model of choice, Amos & I, pub’d 1979) — subseq.
chapters: ways human choices deviate fr. rationality, tendency to treat
problems in isolation, & w. framing effects, decisions shaped by inconsequential
features of choice problems — a deep challenge to the rationality assumption
favored in standard economics
·
Part 5. Two Selves: disting. experiencing
self vs. remembering self, hv diff. interests — e.g. expose people
to 2 painful exper’s, one worse, i.e. longer, but we can exploit the auto. formation
of memories (a feature of S1) so that worse episode leaves better memory; when
people later choose which episode to repeat, they are guided by their rememb.
self, expose selves to unnecess. pain —2 selves theory applied to measurement
of wellbeing, what makes the experiencing self happy is diff. fr. remembering
self — difficult question: How can two selves within a single body pursue
happiness? for individuals, for
societies that view well-being of pop’n as policy objective
·
Concluding chapter: in reverse order,
implications of three distinctions drawn in the book: (1) experiencing vs. remembering
selves, (2) conception of agents in classical economics vs. in behavioral
economics (which borrows fr. Psych’y), (3) auto. System 1 vs. effortful System
2 — then, the virtues of educating gossip, how organizations can improve the
quality of judg’ts & decisions made on their behalf
·
Appendixes: 2 articles author wrote w. Amos
—review of judg’t under uncertainty — prospect theory, studies of framing
effects — cited by the Nobel committee—suprisingly simple —will give you a
sense of how much we knew long ago, also how much learned in recent decades
01 - Chapter 1: The Characters of the Story
Part 1: Two Systems
Chapter 1: The
Characters of the Story
System 1
- S1 operates auto’ly, quickly, little or no effort, involuntary — e.g. see pict. of angry woman, know inst’ly her mood, anticip. her likely to say next
·
e.g.:
o detect that one object is more distant than another
o orient to source of sudden sound
o complete the phrase “bread and…”
o make a “disgust face” when shown horrible picture
o detect hostility in a voice
o answ. to 2 + 2 = ?
o read words on large billboards
o drive car on empty road
o find strong move in chess (if you are a chess master)
o underst. simple sentences
o
recog. that a “meek & tidy
soul w. a passion for detail” resembles an occupa’l stereotype
- S1 incl. innate skills we share w. other animals — e.g. recog. objects, orient attention, avoid losses
- S1 incl. mental activities that become fast & auto. thru prolonged practice — assoc’ns between ideas, learned skills such as reading & underst. nuances of social situations — also expert skills, e.g. chess intuition
- S1 often completely voluntary — but some can be partially controlled, e.g. chewing
System 2
- S2 is effortful, attention req. allocation (“pay attention”) — often assoc’d w. feeling of agency, choice, concentr’n — the conscious, reasoning self, has beliefs, makes choices, decides what to think about & what to do —e.g. multiply 17 x 24
·
e.g.:
o brace for starter gun in a race
o focus on voice of a particular person in a crowded, noisy room
o look for woman w. white hair
o search memory, identify a surprising sound
o maintain a faster walking speed than is natural for you
o monitor appropriateness of your behavior in a social situation
o count the occurrences of letter ‘a’ in a page of text
o tell someone your phone number
o park in narrow space
o compare two washing machines for overall value
o fill out tax form
o check validity of complex logical argu’t
- all S2 operations req. attention, disrupted when attention is withdrawn —effortful activities interfere w. each other, difficult to do several S2 things at once, unless very easy activities — Invisible Gorilla experiment: focused attention à blind to the obvious; also, blind to own blindness
Plot Synopsis (Organization of the book)
- impt. theme: interaction of S1 & S2
- both S1 & S2 are active whenever we are awake — S1 runs auto’ly, S2 normally in low-effort mode, using only fraction of its capacity
- S1 contin’ly generates suggest’ns for S2: impressions, intuitions, intentions, feelings — if endorsed by S2, impressions & intuitions turn into beliefs, impulses turn into voluntary actions — usually S2 adopts the suggestions of S1 w. little or no modification — usually this is efficient (we gen’ly believe our impressions, act on our desires) — but not always
- S2 activates when S1 does not have an answ. — S2 activated for event wh. violates S1’s model of the world — e.g. a surprise event activates & orients your attention: you stare, search your memory for a story that makes sense of the event
- S2 also monitors our own behavior (e.g. stay polite when angry, alert for danger)
- S2 mobilized when error is about to be made — e.g. think of when almost blurted out something, hard to control
- Summary: most of what you (your System 2) think & do originates in your System 1, but System 2 takes over when things get difficult, S2 normally has the last word.
- S1 & S2 usually work well together — however, S1 has biases, leading to systematic errors in specified circumstances — e.g. S1 sometimes answ’s easier q. than the one it was asked, S1 little underst. of logic & statistics
- also, S1 cannot be turned off
Conflict
·
this test engages S2:
Your first task
is to go down both columns, calling out whether each word is printed in
lowercase or in uppercase. When you are done with the first task, go down both
columns again, saying whether each word is printed to the left or to the right of center by saying (or whispering
to yourself) "LEFT" or "RIGHT."
LEFT upper
left lower
right LOWER
RIGHT upper
RIGHT UPPER
left lower
LEFT LOWER
right upper
- conflict between auto. reaction & intention to control — S2 overcoming the impulses of S1 — i.e. S2 is in charge of self-control
Illusions
- Müller-Lyer illusion — S1 “sees” lines are unequal, error — measuring (S2) shows they are equal
- e.g. cognitive illusion: psychotherapist feels strong sympathy for patient w. repeated treatment failures (psychopathic charm)
- N.B. cognitive illusions v. difficult to overcome — S1 automatic, cannot be turned off at will — S2 may have no clue to biases of S1 — cannot be contin’ly vigilant
- the best we can do, compromise: learn to recog. situations in which mistakes are likely, try harder to avoid signif. mistakes when the stakes are high
- N.B. it is easier to recog. other people’s mistakes than our own
Useful Fictions
- S1 & S2 are only personifications, useful fictions
Speaking of System 1 & System 2
·
“He had an impression, but some
of his impressions are illusions.”
·
“This was a pure System 1
response. She reacted to the threat before she recognized it.”
·
“This is your System 1 talking.
Slow down & let your System 2 take control.”
02 - Chapter 2: Attention & Effort
Chapter 2: Attention & Effort
- S2 is lazy, usually guided by S1 — but S2 has vital tasks
Mental Effort
- e.g. strenuous exer. for S2, Add-1 = first, think of strings of 4 digits, all different, write each string on an index card. Place a blank card on top of stack. Add-1, beat steady rhythm (or use metronome, 1/sec). Remove the blank card, read the four digits aloud. Wait for two beats, then report a string w. each of the original digits incremented by 1 — (e.g. if the digits on the card are 5294, report 6305) — impt. to keep rhythm — even more difficult, Add-3.
- pupil dilation = mental effort — window to the soul — cognitive pupillometry
- mental life (S2) normally low level, relaxed — occasionally speeded up, even less often makes big effort
- Difficult mental task àblindness — e.g. Invisible Gorilla experiment — S2 allocates attention to most impt. task, leftover for other tasks, on sec.-by-sec. basis— in an emergency S1 has full control
- more skill involves less effort — “law of least effort,” least demanding course of action
- What makes some cognitive operat’ns more demanding & effortful than others? What req. increased attention? What can S2 do that S1 cannot?
- effort is required to keep 2 ideas in mind at same time, each requiring separate actions, or to combine the two (S2)
- S2 is the only one that can follow rules, compare objects on several attributes, make deliberate choices between options
- S1 detects simple relations (“they are all alike,” “son is much taller than father”), easily integrates information about one thing — but S1 does not deal w. multiple distinct topics at once — S1 not good w. purely statistical information
- S2 impt. capability: adopt “task sets,” i.e. program memory w. an instr’n overriding habitual responses — e.g. count all occurrence of letter f on page — “executive control” = adopt & terminate task sets — effort is increased when switching tasks, also by time pressure
Speaking of Attention & Effort
·
“I won’t try to solve this
while driving. This is a pupil-dilating task. It requires mental effort!”
·
“The law of least effort is
operating here. He will think as little as possible.”
·
“She did not forget about the
meeting. She was completely focused on something else when the meeting was set and
she just didn’t hear you.”
·
“What came quickly to my mind
was an intuition from System 1. I’ll have to start over and search my memory
deliberately.”
03 - Chapter 3: The Lazy Controller
Chapter 3: The Lazy Controller
- S2 has a natural speed — while monitoring envir’t, random thoughts, minor decisions
- self-control is req’d for coherent or effortful thinking, fighting inherent laziness (“law of least effort”)
- sometimes cognitive thinking achieves flow = “state of effortless concentr’n so deep that we lose sense of time, of self, of personal problems” — in a state of flow, no sf-control req’d to maintain focused attention on intense activities, resource of sf-control are freed to be directed to task at hand
The Busy & Depleted System 2
- sf-control (S2) takes mental effort, so when cognitive effort (S2) is high, sf-control is reduced — e.g. when cogn’ly busy, more prone to give in to tempt’n, to S1 choices, selfish choices, sexist lang., superf’l social judg’ts
- S2 controls thought & behaviour
- ∃a shared pool of mental energy, all voluntary effort (cognitive, emotional, physical) draws at least partly on it — any effort of will or self-control is tiring, reduces sf-control —lost motivation, after sf-control in one task, not feel like making an effort in another —called ego depletion — not just a metaphor, consumption of glucose; can restore sf-control by taking glucose — e.g. Israeli judges grant parole more often soon after food breaks
- e.g. activities that deplete self-control — try not think of white bears, stop emotional response to a stirring film, make series of choices involving conflict, try impress others, respond kindly to partner’s bad behavior, interact w. a person of a different race (for prejudiced indiv’ls)
- e.g. indications of depletion —deviate fr. one’s diet, overspend on impulsive purchases, react aggress’ly to provoc’n, persist less time in handgrip, perform poorly in cogn. tasks & logical decision making
The Lazy System 2
- S2 monitors & controls suggestions of S1 — e.g. puzzle: A bat & ball cost $1.10. The bat costs one dollar more than the ball. How much does the ball cost? — intuition (S1) suggests easy, incorrect answer, 10 cents — failure to check answ. w. simple math task, failure to think of social cue, Why asked such an “obvious” question?
- e.g. syllogism: All roses are flowers. Some flowers fade quickly. Therefore some roses fade quickly. — seems intuit’ly valid, but it’s invalid, checking the logic req. S2 — concl’n: if believe a conclusion is true àbelieve argu’ts that seem to support it, even invalid ones — for S1, concl’n comes first, argu’ts follow
- e. g. the Michigan/Detroit problem — How many murders occur in the state of Michigan in one year? — answ. depends on memory that Detroit (high crime) is in Michigan, retrieved more quickly in some people, but can be found w. an S2 search of memory
- these 3 examples show S2 is lazy, accepts easy S1 responses — people less prone to these errors are more intellect’ly engaged, “rational”
Intelligence, Control, Rationality
- sf-control is a measure of intelligence
Speaking of Control
·
“She did not have to struggle
to stay on task for hours. She was in a state of flow.”
·
“His ego was depleted after a
long day of meetings. So he just turned to standard operating procedures
instead of thinking through the problem.”
·
“He didn’t bother to check
whether what he said made sense. Does he usually have a lazy System 2 or was he
unusually tired?”
·
“Unfortunately, she tends to
say the first thing that comes into her mind. She probably also has trouble
delaying gratification. Weak System 2.”
04 - Chapter 4: The Associative Machine
Chapter 4: The Associative Machine
- the surprising workings of S1 — e.g. consider the words “Bananas Vomit” — S1 assumes the words are connected, temporally, causally à slight nausea, some phys’l responses, altered memory ( i.e. more responsive to assoc’d words), sense of surprise
- associative activation — responses are instantaneous, involuntary
- N.B. this complex of responses is coherent — a
self-reinforcing pattern of cognitive, emotional, & phys’l responses
that is both diverse & integrated, associatively coherent
- starting fr.
completely unexpected event (2 simple words, oddly juxtaposed),
- S1 tried
to make sense, linked words in causal story;
- evaluated
possible threat (mild to moderate);
- created
context for future develop’ts, preparing for events suddenly now more
likely;
- also
created a context for the current event, evaluating how surprising it was
- note that
for S1 mere conjunction of 2 words rep’d reality, hence body reacted as
mild reaction to real thing, also emotional response & phys’l recoil included
— cognition is embodied, i.e.you think w. your body, not only brain
- assoc’n of ideas, associative memory —much happens at once, an idea is activated, activates many
ideas, they activate others, often unconscious
- we have limited access to the workings of our minds
The Marvels of Priming
- priming effect — once exposed to one word, more ready to detect assoc’d words — then the primed idea can prime more ideas, altho more weakly, ripple effect
- priming can be result of event (not just words, ideas), we may not be aware — e.g. “the Florida effect,” students in Fla. primed w. words suggesting old age afterwards walked more slowly, i.e. word primed idea of old age, wh. then primes behaviour; no awareness — the ideomotor effect, an idea primes an action — reciprocal priming, walking abnormally slowly primes ideas of old age — nodding head primes for positive response, shaking head for negative
Primes that Guide Us
- priming shows we are not as autonomous as we think
- e.g. voting affected by location of polling booth, booth in school increases support for educ’n — priming w. reminders of money à more independent, persistent, selfish, less social, i.e. individ’ism —images of Big Brother à less independent thinking — reminders of mortality à more open to authoritarian ideas — shame àinclination to think of & buy soap (the “Lady Macbeth” effect)
-
e.g. image placed in front of coffee machine, how many people pay for their coffee? - conscious self does not want to believe in power of priming, but it is true
- S1 is the “stranger in you” — largely unknown, but in control of much of what you do — S1 give impressions à beliefs — S1 impulses à choices, actions — S1 interprets what happens to you & around you, linking the present w. the recent past & w. expect’ns abt the near future — S1 contains the model of the world that instantly evaluates events as normal or surprising —source of your rapid & often precise intuitive judg’ts —does most of this without conscious awareness of its activities
- S1 is also the origin of many of the systematic errors in our intuitions
Speaking of Priming
·
“The sight of all these people
in uniforms does not prime creativity.”
·
“The world makes much less
sense than you think. The coherence comes mostly from the way your mind works.”
·
“They were primed to find
flaws, and this is exactly what they found.”
·
“His System 1 constructed a
story, and his System 2 believed it. It happens to allel
·
“I made myself smile and I’m
actually feeling better!”
05 - Chapter 5: Cognitive Ease
Chapter 5: Cognitive Ease
- mind is always active to maintain & update certain q’s — e.g. Is anything new going on? A threat? Things going well? Shd attention be redirected? Task req. more effort? — these assessments determine (among other things) whether extra effort is req’d fr. S2
the mind constantly measures cognitive ease, i.e. how well things are doing, whether extra effort fr. S2 is needed — cognitive strain
- e.g. sentence printed in clear font, or repeated, or primed, is processed w. cogn. ease — hear speaker when you are in a good mood (or even when pencil stuck crosswise in your mouth to make you “smile”) induces cogn. ease — converse also true for cogn. strain
- causes of ease or strain have interchangeable effects — when at cogn. ease, mood is good, like what you see, believe what you hear, trust intuitions, current situ’n feels comfortably familiar — also thinking likely to be rather casual & superficial — when strained, more vigilant & suspicious, more effortful, less comfortable, make fewer errors — but also less intuitive, less creative
Illusions of Remembering
- illusion of familiarity — e.g. shown fictitious names, days later given list of names, asked to pick out minor celebrities, will include the fictitious names — sense of familiarity for minor items ß sense of ‘pastness,’ an illusion, only because you recog. the word more easily — cognitive ease of S1 (see illustration above) — illusion ß not aware of why cogn’ly easy
Illusions of Truth
- some illusions are predictable when judg’t based on sense of cognitive ease or strain (see illustration above) — beliefs biased by anything that makes it easier for associative machine to run smoothly — freq. repetition à false beliefs, familiarity is confused w. truth
How to Write a Persuasive Message
- reduce cogn. strain — e.g. print w. max contrast betw characters & background, strong colours, high quality paper, simple lang. — memorable ß rhyme, rhythm — if quoting a source, source’s name easy to pronounce
- know a statement is true ß strongly linked by logic or assoc’n to pre-existing beliefs or pref’s, comes fr. trusted source — i.e. sense of cogn. ease — BUT cogn. ease can have many sources, we are not always aware of which
- sense of ease or strain has multiple causes, hard to separate them —but not impossible, req’s strong motiv’n to overcome superf’l factors that produce illusions of truth
- usually lazy S2 adopts the suggestions of S1
Strain & Effort
·
symmetry: S2 ßà cogn. strain — i.e. cogn. strain mobilizes S2
·
e.g. puzzle w. obvious but
wrong answ. dupes people less when printed in small font, strain awakens S2
The Pleasure of Cognitive Ease
- cogn. ease of S1 gives good feelings — e.g. IPOs more successful for companies w. easy to pronounce names — the mere exposure effect: repeating unknown word increases good feelings about it, even unconscious (word shown rapidly) — biology, repeated stimulus, learn ∃ no bad consequences
Ease, Mood, & Intuition
- tests of creativity
- these go together: good mood, intuition, creativity, gullibility, increased reliance on S1
- these go together: sadness, vigilance, suspicion, an analytic approach, increased effort
- happy mood loosens the control of S2 over performance, more intuitive & creative, but less vigilant, more prone to logical errors
- good mood = things gen’ly going well, environ’t is safe, OK to let guard down
- bad mood = things not going very well, may be a threat, vigilance is req’d
Speaking of Cognitive Ease
·
“Let’s not dismiss their
business plan just because the font makes it hard to read.”
·
“We must be inclined to believe
it because it has been repeated so often, but let’s think it through again.”
·
“Familiarity breeds liking.
This is a mere exposure effect.”
·
“I’m in a very good mood today,
and my System 2 is weaker than usual. I should be extra careful.”
Subscribe to:
Posts (Atom)