Description
Tests which relational inferences users carry from a source domain metaphor into a product, revealing correct transfer, missing transfer and harmful over-transfer.
When a product borrows a metaphor (wallet, folder, cart, inbox, assistant), users infer how it behaves by mapping relations from the familiar source domain onto the new target. Structure-mapping theory predicts that people transfer relational structure, not surface features, and that they may also transfer relations the designer never intended. An analogical transfer test presents the metaphor and asks users to predict system behaviour in scenarios that probe specific relations: is money in the wallet insured, can someone take it, does it disappear if the phone is lost? Answers are scored against the system's actual behaviour. Rigour comes from deriving probes systematically from the source-domain relational structure, including probes for intended, unintended and absent mappings, and from comparing alternative metaphors on the same probe set.
Attributes
- Discipline
- Cognitive Psychology & Perception
- Loss type
- Unmigrated
- Stage
- Design, Evaluation
- Data
- Mixed
- Effort
- Days
Steps
- Map the source domain's key relations (for a wallet: carried on the person, can be lost or stolen, holds cash, no interest, no institution behind it) and the target system's actual relations.
- Classify each relation as intended transfer, harmful transfer (true in source, false in system) or missing transfer (true in system, absent in source).
- Write 8 to 15 scenario probes, each targeting one relation, phrased as predictions: 'Your phone is stolen. What happens to the money?'
- Present the metaphor in context to 15 to 30 participants per segment and collect predictions and confidence ratings (1 to 5) for each probe.
- Score each prediction as correct or incorrect against system behaviour and compute per-relation transfer rates, flagging harmful transfers held by more than 20% of participants with high confidence.
- Repeat with one or two alternative metaphors on the same probes and choose the metaphor with the highest correct transfer and lowest confident harmful transfer, or plan explicit explanation for the remaining gaps.
Sources
- Pocket Universal Principles of Design (Lidwell)
- Methods Atlas (Gopinath)
- Embodied Mind (Varela)
- The Inquiry Series, Ch 20 — Semiotics, Icons & Cross-Cultural Sign Failure
Also known as
- Metaphor transfer test
- Structure-mapping test
- Source-target inference test
Categories
- User Research
- Cognitive Psychology & Perception