Attribute Substitution
Expose the easy question your intuition answered instead
- Difficulty
- Easy
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 98%
Attribute substitution explains how intuitive judgment handles a difficult question by silently replacing it with an easier, related one. A person asked about an event's probability may instead assess how surprising or representative it feels; someone asked about overall happiness may report their present mood. Because the substitute answer arrives fluently, the person experiences it as an answer to the original question. The practical method is to name the target question, capture the immediate answer, and then ask what simpler attribute produced it. Decision-makers can then test whether that attribute is valid evidence, add missing information, and deliberately answer the actual question. The framework turns an invisible cognitive shortcut into an inspectable step in reasoning.
Origin
Developed by Daniel Kahneman and Amos Tversky from their research into heuristics and cognitive illusions; Kahneman presented it as the general process underlying several familiar biases.
Core principles
- 01Intuition substitutes tractable questions for difficult ones
- 02A fluent answer is not necessarily an answer to the question asked
- 03Recognizing the substitute question improves judgment
- 04Complex judgments require deliberate checks
How to run it
- 1
Name the target question
Write the difficult question in precise terms, including the outcome and time horizon being judged.
Pro tip Define what a satisfactory answer would measure before considering evidence.
Watch out A vague target makes substitution difficult to detect.
- 2
Capture the intuitive answer
Record the first answer or impression that comes to mind without defending it.
Pro tip Notice whether the answer arrived unusually quickly or confidently.
Watch out Do not mistake fluency for relevance.
- 3
Find the substitute
Ask which easier question your intuition may actually have answered, such as how familiar, surprising, typical, or emotionally vivid something feels.
Pro tip Phrase the substitute as a complete question and compare it directly with the target.
Watch out The substitute can be related to the target while still being an unreliable proxy.
- 4
Test the proxy
Determine whether the substitute attribute has a demonstrated relationship to the target judgment and identify evidence it ignores.
Pro tip Use base rates, comparison classes, and objective measures where available.
Watch out A compelling anecdote does not establish a reliable relationship.
- 5
Reanswer deliberately
Return to the original question and form a revised answer from relevant evidence, explicitly noting uncertainty.
Pro tip Keep the intuitive answer visible so any revision is conscious.
Watch out Systematic biases can persist even after you recognize them.
In the wild
A manager asks an employee how happy they have been at work over the past year. The employee has just left a frustrating meeting and immediately gives a low rating. Applying attribute substitution reveals that the employee answered “How do I feel right now?” The manager separates the recent event from the year-long assessment and reviews workload, relationships, growth, and prior feedback.
→ The discussion produces a more representative assessment and isolates the meeting as a specific problem.
An investor calls a market crash highly probable because a dramatic news event dominates the day's coverage. The investor identifies the substitute question—“Can I easily imagine a crash?”—then checks historical frequencies, valuations, positioning, and competing scenarios before changing the portfolio.
→ The allocation decision reflects probability evidence rather than the vividness of one event.
Common mistakes
Treating every intuition as useless
Attribute substitution does not imply that intuitive judgments are always wrong; it asks whether the intuition answers the intended question in this environment.
Naming a bias without redoing the judgment
Recognizing substitution is only diagnostic. The original question still needs an evidence-based answer.
Is it for you?
Best for
People making judgments about probability, performance, happiness, risk, or other hard-to-measure outcomes.
Not ideal for
Routine decisions where experience provides reliable, rapidly verifiable feedback.
From the transcript
“What I'm thinking of now is that there is a more general process, and that's the process I described a minute ago. They called that…”
“So instead of answering the complex, difficult question, you answer what you can, which is the easier question.”
“what happens is people take the answer to the easy question, they use it to answer the difficult question, and they think they have answered…”
From the episode
Interview With Daniel Kahneman: Masters in Business (Audio)
Daniel Kahneman