dual-process-heuristics-and-cognitive-ease

name	dual-process-heuristics-and-cognitive-ease
description	Audits whether a fast, confident judgment in an analytic product was checked by deliberate reasoning, or instead rests on a substituted question, cognitive ease, or an illusory pattern.
kind	skill
status	ready
provenance	{"principles":["P006","P018","P019","P022","P052","P056","P057","P064","P065","P066","P071","P072","P119","P140","P144","P145","P146","P178","P179","P182","P183","P186"],"claims":["C00125","C00143","C00146","C00170","C00171","C00182","C00183","C00190","C00191","C00197","C00198","C00199"],"evidence":[],"source_anchors":[],"authored_from_digest":"ee59c0fc27e66ba0a2ce84e0c413a8104917036ac481f4f83c3bc63f4cd2a951"}

Dual Process Heuristics And Cognitive Ease

Purpose

Audit whether a fast, confident judgment in an analytic product was actually checked by deliberate reasoning, or whether it rests on an unexamined System-1 shortcut. Under WYSIATI the mind treats whatever evidence is already at hand as the whole story, quietly substitutes an easier question for a harder one, and misreads fluency, familiarity, and a coherent-feeling story for truth — so a hard question can receive a quick, confident answer that was never actually interrogated by slower reasoning. This skill audits that reasoning process itself — whether a System-2 check ran, and what shortcut filled the gap when it did not — not the substantive conclusion the judgment reached.

When to use

• A hard, effortful question receives a quick, confident answer, with no visible sign a slower, second check was ever run.
• The judgment reads as resting on whatever evidence happens to be in hand, with no check on that evidence's reliability or on what is missing from it.
• A causal, skill-based, or "recipe for success" account explains a streaky, extreme, or lucky-looking result — a hot-hand run, a firm's rise or fall, a leader's reputation, or any past success or failure credited to talent or intention rather than chance.
• The write-up credits a conclusion to how familiar, fluent, or instantly coherent it reads, or to a fraction-of-a-second sense that it "fits," rather than to independent verification.
• A probability or risk estimate tracks how easily supporting examples come to mind rather than their actual frequency.
• The judgment was reached in a notable mood, under heavy competing cognitive load, or amid an unnoticed situational cue, and the write-up does not account for it.
• An early, blurry first read of the evidence still appears to be anchoring the current conclusion even though clearer information has since arrived.
• Someone's competence, trustworthiness, or authority is being judged mainly from appearance or demeanor rather than from a track record.
• The write-up leans on a claimed expert "feel" or pattern recognition, and whether it was earned through regular, high-feedback practice is unclear.
• One coherent storyline has absorbed every discrepant detail, with no separate note of what does not fit, or an explanation offered that would equally account for two opposite outcomes.

Procedure

1. Identify the specific judgment under review and the exact question it was meant to answer. Check whether that question was actually answered, or whether a quick urge to answer an easier, related question was acted on instead; reread the original question before judging the answer (P119).
2. Confirm a genuine second, slower pass ran on that judgment rather than accepting the first read. An engaged reviewer checks an intuitive answer that came quickly; a "low checker" who runs with the first idea is impulsive and error-prone, and System 1 is not readily educable, so this check has to be structural, not assumed (P072, P178, P006).
3. Test the judgment against its evidence base: does it treat only the evidence already in hand as the whole story, without weighing that evidence's reliability or registering what is missing? Worthless or untrustworthy evidence must be discounted to the same status as no evidence at all (P022).
4. If the conclusion rests on a causal, skill-based, or "recipe for success" account of an extreme, streaky, or lucky-looking result, check whether the underlying group or sample is simply small, whether the opposite tail was tested, and whether the account would have made the outcome predictable in advance rather than only explicable after the fact (P056, P057, P066).
5. Separate warranted intuition from unwarranted confidence: check whether a claimed "feel" or pattern recognition was built in a domain with regular practice and clear feedback, since that is what licenses trusting a fast read; that most cognition is unconscious is not, by itself, evidence the fast read is accurate (P065, P182).
6. Discount familiarity, fluency, and a smoothly coherent story as evidence of truth. A claim that merely feels recognizable, well-fitting, or free of surprise — assessed in a fraction of a second — has not thereby been verified; ask what independent check was actually run (P144, P179, P186).
7. When a probability or risk estimate is offered, check whether it tracks how easily supporting instances came to mind — through fame, salience, or ease of imagining a scenario — rather than their true, checkable frequency (P018).
8. Note any situational factor that could have shaped the judgment outside the reasoner's own awareness: a notable mood (which raises either credulity or vigilance), heavy competing cognitive demand, or an unrelated environmental cue; flag when the write-up does not account for it (P140, P071, P064).
9. Check whether an early, blurred, or ambiguous first read of the evidence is still shaping the current conclusion. The earlier and longer that impression was held, the clearer the later evidence needs to be before the conclusion should be considered updated (P183).
10. If a judgment about a person's competence, trustworthiness, or authority rests mainly on appearance or demeanor, flag it regardless of how widely the impression is shared — such impressions show high agreement but low accuracy (P146).
11. Check whether a single interpretation was adopted without separately registering the discrepant details that do not fit it, or whether the explanation offered would equally account for two opposite outcomes; flag both as unearned coherence rather than confirmed understanding (P019, P145).

Inputs

• The specific judgment, estimate, or conclusion under review, and the exact question it was meant to answer.
• The evidence the analyst had in hand when the judgment was formed, including any known gaps or reliability concerns.
• Any causal, skill-based, or "recipe for success" account offered for an extreme, streaky, or lucky-looking result.
• Context on how the judgment was reached — how quickly, in what mood or under what workload, and whether a second, slower pass was run.
• The basis of any claimed expert "feel" or pattern recognition, including whether it was built through regular, high-feedback practice.
• The sequence in which information arrived, when an early or ambiguous first read may still be shaping the current view.
• Any appearance- or demeanor-based impression feeding a judgment about a person.

Output

Per finding: name the bias or perceptual mechanism (attribute substitution, cognitive ease misread as truth, an illusory pattern in chance, an appearance-driven trait judgment, and so on), apply the principle(s) by ID, give the corrective technique (reread the actual question, discount fluency and coherence as evidence, check the sample size and the opposite tail, weigh the track record behind a claimed intuition), state the residual uncertainty the correction leaves, and end with a concrete next step. Order findings highest-impact first. This skill critiques how the judgment was reached; it never supplies the substantive conclusion in its place.

Anti-patterns to flag

• A hard question answered instantly and confidently, with the easier substituted question never surfaced or reread (P119).
• Skipping the second, slower check because the first answer felt obviously right — the mark of a low-checking, error-prone process rather than a careful one (P072, P178, P006).
• A conclusion defended as sound because "all the evidence points that way," with no check on that evidence's reliability or on what is missing (P022).
• An extreme or streaky result explained by skill, talent, or a "recipe for success," with no check on sample size, the opposite tail, or whether the story would have predicted the outcome in advance (P056, P057, P066).
• A claim accepted because it reads well, feels familiar, or resolves instantly as coherent, with fluency or a fraction-of-a-second sense of fit standing in for verification (P144, P179, P186).
• A probability or risk estimate that tracks how vividly or easily examples come to mind rather than their actual frequency (P018).
• A judgment's mood, workload, or incidental surrounding context never screened for, despite plausibly shaping how confidently or credulously it was reached (P064, P071, P140).
• An early, blurry impression still driving the conclusion, with no acknowledgment that later, clearer evidence should have required more before being accepted (P183).
• A person's competence or trustworthiness inferred mainly from appearance or demeanor, with wide agreement on the impression mistaken for its accuracy (P146).
• A "gut feel" or pattern-recognition claim accepted as expertise with no check on whether it was earned in a regular, high-feedback domain (P065, P182).
• A single tidy narrative that accounts for all the evidence with no separate note of the details that do not fit, or an explanation that would equally justify two contradictory outcomes (P019, P145).
• The bias waved off because the analyst is experienced, incentivized to be accurate, or feels internally consistent — none of which defends against a heuristic bias from the inside (P052).

References

See ../../references/bias-perception-principles-index.md for the full principle catalogue. For adjacent concerns, see the sibling skills: judgment-anchoring-and-base-rates owns the classic anchoring, base-rate-neglect, and representativeness heuristics for judgments under uncertainty; this skill owns the System 1/System 2 architecture behind them, including WYSIATI, cognitive ease, and the availability-by-ease-of-recall heuristic. motivated-reasoning-and-belief-perseverance owns identity-linked belief freezing and discrepant evidence explained away; this skill's coherence-seeking is the automatic perceptual mechanism, not a motivated defense of a preferred conclusion. mind-sets-and-structured-techniques owns applying a structured technique to break an established mind-set once one is identified; this skill flags the underlying System-1 sense-making and pattern-in-chance illusions that form a mind-set in the first place.

Provenance

Derived solely from P006, P018, P019, P022, P052, P056, P057, P064, P065, P066, P071, P072, P119, P140, P144, P145, P146, P178, P179, P182, P183, and P186 (Thinking, Fast and Slow; Psychology of Intelligence Analysis; Superforecasting — all distillation-only; see the frontmatter above for the full claim, evidence, and source-anchor list).