Representativeness Heuristic
Representativeness judges likelihood through resemblance to a typical case, which can neglect base rates and other evidence. The excerpt outlines Study 2 and its matching tasks but does not supply the results needed to assess participants’ errors.
THE IDEA
The representativeness heuristic is a mental shortcut used when making judgments about the probability of an event under uncertainty. People judge the likelihood that an object belongs to a category based on how similar (representative) the object is to the category, often ignoring statistical base rates. In a series of studies by Kahneman and Tversky, participants were given brief personality descriptions and asked to predict the most likely category.
Study 2
The second study had three purposes. In ascending order of importance these were (a) to extend and replicate the earlier study with new and more powerful stimuli; (b) to explore, in addition to the criteria of probability and willingness-to-bet, a third criterion for ranking options that is related to these two but is not normatively obligated to coincide with them; and (c) to directly test some of the speculations set forth to account for the results of the first study.
Stimuli
In Study 2, five new stimuli were used:
- Yossi: A brief personality description of a student matched against areas of university study.
- Doron: A brief personality description matched against sports.
- Dorit and Alon: A description of a place matched against countries or continents.
- Asher: A brief personality description matched against Israeli political parties.
The core mechanism of the heuristic is that people substitute a difficult question ('What is the probability?') with an easier one ('How representative is it?'). This can produce errors such as the conjunction fallacy and neglect of base rates. The excerpt describes the study’s aims and stimulus categories, but does not include the full descriptions, ranking criteria, or results.
When to use it
Investing: A venture capitalist judges a startup's potential success based entirely on the charisma of the founder and a compelling pitch ('this is the next Facebook'), ignoring the high base rate of startup failures.
Talent Management: A hiring manager selects a candidate over others because their background and manner perfectly match the 'ideal company culture' narrative, neglecting the base rate of successful hires from that specific background.
Medical Diagnosis: A doctor immediately suspects a rare genetic disease in a patient because their symptoms perfectly match a textbook case, ignoring the base rate probability of a much more common environmental disease with similar symptoms.
Marketing: A consumer buys a product because the advertising creates a very strong 'representative' image of a happy, successful user, ignoring the product's actual statistical performance or quality ratings.
FROM THEORY TO EVERYDAY LIFE
In practice
Yossi (University Study)
Yossi’s personality description is matched against fields of university study. The task asks participants to rank possible fields; assessing those rankings would require the description, relevant base rates, and study results.
Doron (Sports)
Doron’s personality description is matched against sports. The task separates a judgment of resemblance to a typical player from a judgment about which sport he is likely to play.
Dorit and Alon (Geography)
The Dorit and Alon stimuli match descriptions of places against countries or continents. They extend the matching task beyond judgments about individual people.
Asher (Political Parties)
Asher’s personality description is matched against Israeli political parties. Comparing probability and resemblance rankings could test whether participants treat the two questions alike.
A MOMENT TO REFLECT
Put it to work.
What additional information about the ranking criteria, stimuli, and results would you need to decide whether the study demonstrates base-rate neglect?
How might the rankings in the Yossi and Asher examples change if participants had first been asked explicitly to estimate the base rate of students in each academic department or voters for each political party?
Can you recall a time you or someone you trusted was confident in a binary prediction (e.g., a new hire will succeed, a stock will rise) based on a compelling 'representative' narrative, only for the outcome to be governed more by base rates? How can you build a fallback mechanism to counteract this heuristic?
Often confused with
References
- Subjective probability: A judgment of representativeness — Daniel Kahneman and Amos Tversky (1972)
- Judgment under uncertainty: Heuristics and biases — Amos Tversky and Daniel Kahneman (1974)
CONTINUE THE CONVERSATION
Source texts
- Heuristics and Biases The Psychology of Intuitive Judgment (Thomas Gilovich, Dale Griffin, Daniel Kahneman)