🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!
representativeness_heuristic
The tendency to judge the probability of an event by how similar it is to a prototype or stereotype, rather than by actual statistical likelihood. People substitute the question 'How probable is this?' with 'How similar is this to my mental model?' This heuristic is efficient but leads to systematic errors when similarity and probability diverge.
A person described as 'quiet, organized, and detail-oriented' is judged more likely to be a librarian than a salesperson, despite salespeople vastly outnumbering librarians. The description matches the librarian stereotype, overriding base rate information.
A hiring manager receives two resumes: one from a candidate who attended an Ivy League university and lists chess club and classical piano, and one from a state school graduate with extensive retail work experience. For a data analyst role, she instinctively favors the first, matching him to her mental image of an 'analytical type,' despite the second candidate's directly relevant skills.
When a tall, broad-shouldered man at a party mentions he works in sports, everyone assumes he is an athlete. In reality, he is a sports accountant — but his appearance so closely matches the prototype of 'athlete' that people never consider the far more common sports-adjacent professions.
Binary (yes/no) questions an LLM must answer to identify this aspect:
Is probability being judged by resemblance to a stereotype rather than by data?
Type: binaryAre base rates being ignored in favor of how typical something looks?
Type: binaryWould the judgment change if the assessment were based on statistics rather than similarity?
Type: binaryThe tendency to judge the probability of an event by how similar it is to a prototype or stereotype, rather than by actual statistical likelihood. People substitute the question 'How probable is this?' with 'How similar is this to my mental model?' This heuristic is efficient but leads to systematic errors when similarity and probability diverge.
The brain uses similarity as a quick proxy for probability because it is computationally cheaper. In many everyday situations, representative instances are more probable, but this breaks down when base rates are extreme.
Separate the question 'Does this match my stereotype?' from 'How probable is this given base rates?' Explicitly consider base rates and alternative explanations before judging probability.
This heuristic affects hiring (judging candidates by how they 'look the part'), medical diagnosis (matching symptoms to prototypical diseases), criminal profiling, and investment decisions.
Use these tools to detect, analyze, or train this aspect.