Thinking 2011

Thinking, Fast and Slow

Daniel Kahneman

General, the track now showing: #11
ADHD: #42

The reference work on how human judgement fails, from the man whose research created the field; loss aversion, anchoring, availability and the outside view remain solid and are the vocabulary everything else in this domain uses. Rereads well because it is a catalogue rather than an argument.

The chapter guide

Most of what you believe is produced by a fast, associative, overconfident system that builds coherent stories from whatever information is at hand, while the slow deliberate system that could check it is lazy and usually just signs off.

Ch 1–2: The Characters of the Story / Attention and Effort (System 1 and System 2, pupil dilation, the invisible gorilla)

Introduces the two-system metaphor and shows that deliberate thought consumes a measurable, limited resource.
Ch 3: The Lazy Controller (ego depletion, the glucose-and-willpower studies, flow)

Self-control draws down a depletable resource, and eating restores it.
Ch 4: The Associative Machine (the Florida effect, money priming, the Lady Macbeth effect)

Words and concepts unconsciously prime behaviour, so reading ‘Florida’ makes you walk more slowly.
Ch 5: Cognitive Ease (fluency, mere exposure, the illusion of truth)

Things that are easy to process feel more true, more familiar, and more likeable.
Ch 6–7: Norms, Surprises and Causes / A Machine for Jumping to Conclusions (halo effect, confirmation bias, WYSIATI)

The mind builds a coherent causal story from whatever it has and never registers what is missing. What You See Is All There Is.
Ch 8–9: How Judgments Happen / Answering an Easier Question (intensity matching, substitution, the mood heuristic)

When faced with a hard question the mind silently answers an easier one and reports the answer as if it addressed the original.
Ch 10: The Law of Small Numbers

Small samples produce extreme results, and both laypeople and researchers systematically over-read them.
Ch 11: Anchors

Arbitrary numbers contaminate subsequent numerical judgments, with a measurable anchoring index.
Ch 12–13: The Science of Availability / Availability, Emotion and Risk (availability cascades, affect heuristic)

You judge frequency by how easily examples come to mind, so vivid and recent events distort risk perception.
Ch 14–15: Tom W’s Specialty / Linda: Less Is More (base-rate neglect, the conjunction fallacy)

Representativeness overwhelms base rates, and people rate a detailed story as more probable than a subset of it.
Ch 16–17: Causes Trump Statistics / Regression to the Mean

Statistical base rates are ignored unless made causal, and unusual outcomes drift toward average for purely mathematical reasons.
Ch 18: Taming Intuitive Predictions

A formal procedure for shrinking intuitive forecasts back toward the base rate.
Ch 19–20: The Illusion of Understanding / The Illusion of Validity (narrative fallacy, hindsight bias, stock pickers)

Coherent stories about the past create false confidence about the future, and subjective confidence tracks story quality rather than accuracy.
Ch 21–22: Intuitions vs Formulas / Expert Intuition: When Can We Trust It?

Simple algorithms beat expert judgment in low-validity environments, and intuition is trustworthy only where the environment is regular and feedback is fast.
Ch 23: The Outside View (the planning fallacy, reference class forecasting, the textbook that took eight years)

Ask how long this class of project usually takes rather than how long this project will take, because the inside view is systematically and hugely optimistic.
Ch 24: The Engine of Capitalism (optimistic bias, entrepreneurial overconfidence, the premortem)

Optimism drives risk-taking and mostly punishes the optimist, and the premortem is a cheap institutional fix.
Ch 25–28: Bernoulli’s Errors, Prospect Theory, The Endowment Effect, Bad Events (loss aversion, reference points, the negativity dominance)

Value is assessed as gains and losses from a reference point, and losses loom roughly twice as large as equivalent gains.
Ch 29–32: The Fourfold Pattern, Rare Events, Risk Policies, Keeping Score (mental accounting, the sunk-cost fallacy, the disposition effect, regret)

Small probabilities are overweighted, gains and losses are tracked in separate mental accounts, and people throw good resources after bad to avoid closing an account at a loss.
Ch 33–34: Reversals / Frames and Reality (joint vs separate evaluation, the Asian disease problem, MPG illusion)

The same option is valued differently depending on how it is described and what it is evaluated against.
Ch 35–38: Two Selves, Life as a Story, Experiencing Well-Being, Thinking About Life (the cold-hand experiment, the peak-end rule, duration neglect, the focusing illusion)

The experiencing self and the remembering self disagree, and you systematically choose in favour of the one that keeps the memories.
Appendices A and B (the two original Science papers)

Reprints of the 1974 judgment-under-uncertainty and 1979 prospect-theory papers.
Structure and reading strategy (499 pages, 38 chapters, ‘Speaking of…’ recap boxes)

How to actually get through it.

Evidence check

Reliability 4 of 5

Social/behavioral priming: elderly-words make you walk slower (Bargh), warm cup makes you judge people as warmer (Williams & Bargh), Florida effect, money priming. Chapter 4: “disbelief is not an option… you have no choice but to accept that the major conclusions of these studies are true.”

failed replication: the priming literature is the epicenter of the replication crisis. Ulrich Schimmack’s R-Index analysis (“Reconstruction of a Train Wreck”, 2017) found Chapter 4’s cited studies have very low replicability. Kahneman himself wrote an open letter in 2012 calling social priming “a train wreck looming”, and in a 2017 comment on Schimmack’s post conceded he had “placed too much faith in underpowered studies” and that the chapter’s confident wording was a mistake. This is the strongest possible form of author concession.
Ego depletion / willpower as a depletable resource (and the glucose variant).

failed replication: Hagger et al. 2016 multilab preregistered replication (23 labs, N=2141) found an effect near zero; Vohs et al. 2021 multisite paradigmatic test, co-authored by the original theorist, again found no reliable effect. Baumeister and Vohs contest the manipulations used, so the literature is not formally settled, but there is no dependable ego-depletion effect to build advice on. Glucose-restores-willpower is effectively dead.
Facial-feedback (pen-in-mouth), ‘Lady Macbeth’ cleansing effects, and other one-study social-psych vignettes.

failed or weakened replication: the Wagenmakers et al. 2016 RRR failed to replicate facial feedback; several similar chapter anecdotes are underpowered single studies.
Core framework: System 1/System 2, anchoring, loss aversion, availability, framing, prospect theory, base-rate neglect, peak-end and the experiencing/remembering self.

replicated / robust: anchoring and framing are among the most replicable effects in psychology; prospect theory is a Nobel-cited, repeatedly replicated result. Loss aversion has been contested at the margins (David Gal & Derek Rucker argue it is overstated and context-dependent) but the central bias catalogue stands.

Yes. This is the strongest survivor on the list and a legitimate prefix candidate. Roughly one chapter (priming) and one concept (ego depletion) are dead; the architecture and the major biases are intact, and the author publicly corrected himself, which is exactly the epistemic behavior you want to model. Caveat to attach: ‘Chapter 4 on priming, and the ego-depletion material, did not survive replication; Kahneman conceded this.’

Scorecard

Universality: 4 of 5
Evidence: 3 of 5
Reread value: 4 of 5