Description
Asks users to predict what interface elements will do before using them, measuring where judgements based on resemblance to familiar patterns produce mistaken expectations.
The representativeness heuristic describes how people judge likelihood by similarity to a prototype rather than by base rates or underlying logic. Applied to interfaces, users judge what a control does, or whether a feature will work, by how much it resembles patterns they already know. When an element looks like a familiar pattern but behaves differently, users form confident but wrong expectations, producing mistake-type errors (wrong intention) rather than slips. The audit makes this measurable: participants predict an element's behaviour before interacting, and the gap between prediction and actual behaviour is coded. Elements with high prediction-gap rates are flagged as misleading prototypes. Rigour comes from collecting predictions before any exposure, using a coding scheme for gap size, and setting explicit flag thresholds such as more than 30% of participants with a substantial gap.
Attributes
- Discipline
- Behavioral Science & Economics
- Loss type
- Latent
- Stage
- Design, Evaluation
- Data
- Mixed
- Effort
- Days
Steps
- Present users with interface elements, ask them to predict behavior before interaction.
- Code gap between prediction and actual behavior.
- Flag elements where > 30% of participants show prediction gap > 50%.
Sources
- Heuristics and Biases — Kahneman (ed.)
- Tversky & Kahneman, Judgment under Uncertainty: Heuristics and Biases (Science, 1974)
Also known as
- Prototype-match prediction audit
- Affordance prediction gap test
Categories
- Analytics & Behaviour
- Behavioral Science & Economics