frameworks / cognitive-biases-and-psychology · article
frameworks/ · 2,474 w · 11 min · ✎ dialogue

Cognitive Biases and Psychology

The research foundation underlying Jason's coaching work, writing, and personal philosophy — spanning cognitive psychology, social psychology, and neuroscience. Human cognition is not a neutral recording device. It is an evolved prediction machine that generates plausible narratives from incomplete data, systematically favors ease over accuracy, carries biases absorbed from the surrounding culture before conscious thought can intervene, and responds to errors and rewards in ways that differ fundamentally from person to person. Understanding these mechanisms is not an academic exercise — it determines how we coach, how we write, and how we make consequential decisions.

Kahneman's Two Systems: Fast and Slow Thinking

Kahneman's central insight, developed across decades of research with Amos Tversky, is that human cognition runs on two parallel systems with very different operating characteristics.

System 1 is automatic, fast, effortless, and always on. It recognizes faces, detects emotion in voices, jumps to stereotyped conclusions, and reads large text on billboards without any conscious decision to do so. It is gullible and biased toward belief — when we hear a statement, System 1 begins by treating it as true and flags it for review only if System 2 intervenes.

System 2 is deliberate, slow, effortful, and limited. It does multiplication, searches a crowd for a specific person, monitors behavior in a complex social situation. Crucially, it taxes a finite resource: pupil dilation studies show that intense System 2 work literally increases heart rate. As the load increases, performance degrades — and with it, the ability to resist temptation or maintain self-control.

The practical consequence: System 2 thinks it is in charge, but most of the action originates with System 1. System 2 often functions less as a critic and more as an apologist — rationalizing decisions that System 1 had already made. Kahneman calls the resulting failure mode WYSIATI (What You See Is All There Is): we construct confident, coherent narratives from whatever information is currently available, without noticing what is absent. A juror who hears only the prosecution's case will be more confident in her verdict than one who hears both sides, because a consistent one-sided story feels more compelling than a more complete but messier picture.

Key biases arising from this architecture:

  • Anchoring: The first number you encounter distorts all subsequent estimates, even when the anchor is random. In one study, participants who spun a rigged wheel landing on 65 before estimating the percentage of African nations in the UN gave significantly higher estimates than those whose wheel landed on 10. The anchor contaminates without awareness.
  • Availability/Recallability: We judge frequency and probability by ease of recall. Plane crashes feel more common than car crashes because they are more vivid, not because they are more frequent. This systematically skews risk assessment.
  • Confirmation bias: System 2 tests hypotheses by searching for positive evidence, not by trying to falsify them. We seek data that confirms what we already believe and discount what challenges it.
  • Status quo bias: People overwhelmingly stick with defaults even when alternatives are demonstrably better. The default is interpreted as a recommendation — changing it requires expenditure of scarce cognitive resources.
  • Sunk cost fallacy: Continued investment based on past spending rather than future value. The rational response — treat past costs as gone — violates the narrative coherence System 1 demands.
  • Overconfidence: We systematically overestimate our knowledge and the accuracy of our predictions. Experts are often more overconfident than novices because they have developed fluent narratives.

Priming: The Environment Shapes Behavior Below Awareness

Priming research demonstrates that ambient stimuli alter cognition and behavior without any conscious awareness of the influence. This is among the most counterintuitive findings in psychology because it directly contradicts the sense that we control our own behavior.

Classic demonstrations: People who unscrambled sentences containing words like Florida, gray, bingo, wrinkle walked more slowly down the hall afterward than control participants — their behavior literally aged. The reverse held too: people forced to walk slowly for 30 minutes were subsequently faster at recognizing words associated with old age. Participants who held a pencil horizontally in their teeth (forcing a Duchenne smile) rated cartoons as funnier. Arizona voters polled in a school were more likely to support education funding than those polled elsewhere. People made to think about an action they were ashamed of later chose wash and soap over wish and soup to complete word fragments — as if trying to cleanse a metaphorically stained soul.

Cognitive ease is a related mechanism. We tend to believe things that are easy to process and distrust things that require effort. Repeated exposure makes claims feel familiar, and familiarity is misread as truth. Rhyming aphorisms feel more insightful than unrhymed equivalents with identical content. Names that are easy to pronounce are judged as more credible. Bold, high-contrast text is more persuasive. The troubling implication: falsehoods repeated enough times eventually feel true, not because we have evaluated them but because they have become cognitively fluent.

Question order matters in surveys: asking about happiness first, then number of dates last month, yields no correlation. Reverse the order, and the correlation becomes substantial — recent dating experience primes the mood that drives the happiness rating.

Implicit Bias: What You Don't Know About Your Own Prejudices

From 2.5 million+ completions of the Implicit Association Test (Nosek, Banaji, Greenwald, Project Implicit):

Implicit biases operate below conscious awareness. People cannot introspect them reliably — in fact, self-report of racial attitudes is essentially uncorrelated with IAT scores. Approximately 70% of white Americans show implicit pro-white bias on the IAT despite reporting no racial preference. Biases of this kind are pervasive across all groups and topics: race, gender, sexuality, weight, disability, age.

Critically, IAT scores predict behavior. Physicians with higher implicit race bias are measurably less likely to recommend appropriate treatment for Black patients. In hiring contexts, identical resumes with names associated with different racial groups receive significantly different callback rates. Bias doesn't stay abstract — it propagates into consequential decisions.

Dolly Chugh frames implicit bias usefully: it is a habit that formed from the moment you began perceiving the world. Every image on television, every parental comment, every neighborhood norm — these create associations that idle in the background without requiring endorsement. The key property is that the bias is "sticky and malleable": it fluctuates by day and situation, heightened under time pressure and stress. Under cognitive load, System 2 is occupied, and System 1's unexamined associations run the show more freely.

Awareness alone doesn't fix it. Recognizing you have a bias does not remove it. Behavioral interventions that change structure — who gets evaluated together, whether criteria are defined before candidates are reviewed, how networks are curated — are more effective than consciousness-raising alone. The Google executive Rick Klau, after seeing his IAT results showing strong gender bias, wrote code to analyze his professional contacts: 80% were male. His response was structural: he deliberately diversified his network, stopped attending conferences that were 90% male, and tracked his progress. Bias management is an ongoing system problem, not a one-time mindset shift.

Stereotype Content Model: Warmth and Competence

From Fiske, Cuddy, and Glick (originally 2002; extensively replicated):

Social perception organizes around two fundamental dimensions: warmth (intent — is this person/group a friend or foe?) and competence (ability — can they act on that intent?). These two axes are not independent: status predicts perceived competence, and perceived competition predicts low warmth.

This creates four quadrants with distinct emotional and behavioral consequences:

  • High warmth, high competence (mainstream in-group, close allies): elicits admiration; people actively help this group.
  • High warmth, low competence (elderly, people with disabilities, some domestic workers): elicits pity; people passively help.
  • Low warmth, low competence (homeless people, "welfare recipients," undocumented immigrants in many contexts): elicits contempt; people actively harm or ignore.
  • Low warmth, high competence (Asian Americans, Jews, rich professionals): elicits envy — a mixture of admiration and resentment. This is the most dangerous quadrant.

The envy cluster is dangerous because resentment of a competent out-group combines with an attribution that their success comes at your expense. Historically this has driven scapegoating during periods of social stress — Jews in 1930s Germany were positioned in exactly the envious quadrant before violence escalated. The BIAS Map (Cuddy, Fiske, Glick 2007) documents the behavioral outputs: envied groups are targets of passive harm (not helping in a crisis) and, in extreme cases, active violence when threat is salient.

The Asian American case is precisely this model. Lin, Kwan, Cheung, and Fiske (2005) validated the Scale of Anti-Asian American Stereotypes (SAAAS) showing that stereotypes of Asian Americans load on two dimensions: excessive competence and deficient sociability. Six studies showed this envious prejudice predicts limited outgroup friendships and overestimation of Asian presence on campuses. The critical finding: it is the perceived low sociability — not the high competence — that drives rejection. Competence alone would produce admiration. It is the cold intent that produces envy and exclusion.

This helps explain the Harvard admissions pattern revealed by the 2019 lawsuit: Asian applicants scored at or above other groups on academic metrics and scored equally well on alumni interviews, yet received lower "personal ratings" from admissions officers who had never met them. The SCM predicts exactly this: perceived competence without warmth triggers a specific form of prejudice that manifests as social rejection, not academic dismissal.

For anyone navigating this dynamic — including in business or leadership contexts — Fiske's interview note is relevant: "slow rolling" accomplishments to avoid triggering envy is a real and rational strategy in low-stakes social contexts, but creates its own costs in high-stakes situations like fundraising or promotion where you must impress immediately.

See asian-american-leadership for the full leadership implications.

National Character Stereotypes: Agreed Upon, Mostly Wrong

From McCrae et al. (PNAS, 2005):

National character stereotypes — "the Swiss are punctual," "Italians are passionate" — are highly reliable in the sense that people within and across cultures agree on them. But they are virtually uncorrelated with actual measured personality trait distributions in those populations.

The sharpest example: the US and Canada have nearly identical measured personality profiles on the Big Five, yet dramatically divergent national character perceptions. "Americans are outgoing and aggressive; Canadians are polite and deferential" is a story, not data. Stereotypes of national character are social constructions, shaped by national myths, institutional contrasts, historical narratives, and media — not by accumulated accurate observation of individuals.

Gender stereotypes occupy different territory: unlike national stereotypes, they do have a weak "kernel of truth" — measured trait differences, though smaller than stereotypes suggest, exist in the direction stereotypes claim. The practical implication is that gender stereotypes are self-reinforcing in ways national ones are not — they are anchored in a real underlying signal, even if they dramatically exaggerate it.

Error Processing and Learning Styles: The ERN

From Frank and Curran (Neuron, 2005):

The brain does not process errors uniformly. The Error-Related Negativity (ERN) is an electrical signal generated in the anterior cingulate cortex within 100 milliseconds of making a mistake. Crucially, the amplitude of this signal varies systematically across individuals — and that variation predicts how people learn.

People reliably split into two types:

  • Positive learners respond more strongly to reward signals. Dopamine surges after correct choices drive approach behavior and skill acquisition. They learn better from what works.
  • Negative learners have larger ERNs — more neurological activation after errors. Dopamine dips after mistakes support avoidance learning. They learn better from what fails.

This is trait-like, not situational — it persists across contexts and shapes risk tolerance, decision-making style, and emotional responses to criticism. It is not a flaw in either direction; it is a dimension of cognitive architecture.

The coaching implication is direct: people differ fundamentally in whether they are oriented toward seeking reward or preventing punishment. A coach who exclusively uses positive reinforcement will be less effective with strong negative learners, and vice versa. Framing feedback as "here's what to do more" vs. "here's the mistake to avoid" has different impacts on these two populations. Diagnosis precedes prescription.

The Pride System

From Jessica Tracy (Pride, 2016):

Pride is not vanity. Tracy's argument, supported by cross-cultural research including blind athletes, is that pride is the primary motivational engine behind human achievement and cultural innovation. "The desire to feel pride is one of the most important motivational forces propelling human achievement, creation, and innovation, and, as a result, all cultural inventions."

Pride splits into two distinct forms with radically different causes and consequences:

  • Authentic pride arises from effort and hard work — from attributions to controllable, unstable causes. It motivates continued effort, fosters empathy, and builds prestige-based status. People experiencing authentic pride work harder, behave more morally, and function as better leaders.
  • Hubristic pride arises from attributions to stable, uncontrollable traits — "I'm naturally gifted." It is associated with narcissism, low implicit self-esteem, dominance-based status, and disagreeableness. Hubristic pride does not reliably motivate more effort, but does motivate performance in competitive or anger-inducing contexts.

Tracy's research on social learning links pride to cultural evolution: humans are designed to copy the most prestigious members of their groups, and prestige signals competence-with-generosity, not dominance-with-threat. The "looking-glass self" (Cooley, 1902) is the underlying mechanism: we construct identity through the reflected appraisals of others, and pride is the emotional system that makes us care enough about those appraisals to change.

The display calibration problem: Pride displays must be precise. Too little, and status isn't communicated. Too much — arms raised, chest puffed, over-the-top — and observers find it off-putting even when the pride is earned. Winners who showed intense pride expressions were rated as less likable than winners who showed the same emotion more subtly. The norm against arrogance is a social-control mechanism for preventing status rivalries from escalating into constant conflict.

Pride's effects extend to intergroup attitudes: participants primed with hubristic pride rated Asian Americans more negatively; those primed with authentic pride rated them more positively. The type of pride you are experiencing in the moment shapes who you perceive as deserving.

Thread · 0 replies+ add reply
no replies yet — be the first to write back to this article.