This is a Lexicon entry: the mechanism, when it shows up, and the countermeasure. The full correction, with real-world cases and the audit prompt, hasn't been written yet.
Definition
A mental shortcut used when making judgments about the probability of an event based on how much it resembles a typical case. A tall person "looks like" a basketball player so we assume they are one.
When it shows up
Judging how likely someone is to be a librarian versus a farmer from a personality description. Predicting a stock's future from how much it 'looks like' past winners. Any category-guess where a prototype is easier to access than a base rate. Modern context: A recruiter reads a CV from a candidate who went to a prestigious university, had a linear career progression and uses exactly the vocabulary of the job description - and rates them as highly likely to succeed. A candidate with an unconventional path and identical actual competence is rated lower, because they do not match the prototype. High-stakes instance: Hiring and promotion - documented studies show that physical height, voice pitch and name-origin produce systematic effects on hiring and promotion decisions in domains where these variables have no functional relevance. Failure mode: Probability is estimated by resemblance to a mental prototype rather than by frequency. A shy, bookish person gets classified as 'librarian' even when farmers outnumber librarians 50:1 in the relevant population. The base rate is ignored because the prototype match is vivid. Estimated cost: High - often irreversible where this bias dominates unchecked decisions.
Failure mode
Probability is estimated by resemblance to a mental prototype rather than by frequency. A shy, bookish person gets classified as 'librarian' even when farmers outnumber librarians 50:1 in the relevant population. The base rate is ignored because the prototype match is vivid.
Countermeasure
Before using 'looks like' to classify, ask: 'How many of each category exist?' The base rate often swamps the resemblance signal. When correction costs more than the bias: When the representative example is statistically typical and the base rate correction would produce a less accurate answer. Representativeness is a fast, frugal heuristic that works well in stable, well-understood categories. It fails specifically when base rates are very skewed and the representativeness cue is uninformative - not as a general rule.
Related concepts
Also connected in the mapAttribute Substitution·Clustering Illusion·The Naxalt Fallacy·The Base Rate Fallacy·The Conjunction Fallacy·Prototype Effects
Origin
Also revealed
Entry #46 of 631 in The Lexicon · see the full Lexicon