The gulf is the distance between system state and user understanding
Aliases: evaluation gulf · state comprehension · result interpretation
What it is
A Gulf of Evaluation exists when the system has entered a state, but the interface does not transform that state into information users can perceive, interpret, and compare with their goal. Users are not unable to act; they do not know what happened, what it means, or whether the intention was achieved.
Why it happens
Moving from state to understanding requires three conversions: the state must be presented (perception), the presentation mapped to user concepts (interpretation), and the interpretation must support goal comparison (judgment). Asynchrony, permission, terminology, aggregation rules, time basis, and units widen the gulf. Users respond by repeating operations, waiting, screenshotting for proof, or assuming failure.
Studying it
Pause or interview at critical task points and ask users to state current status, the result of the last action, the next step, and completion; compare answers with the true system state. Measure confirmation time, repeated clicks, refreshes, support queries, and wrong judgments. Manipulate feedback visibility, terminology, granularity, and delay to observe interpretation accuracy.
Where it stops holding
More data does not automatically narrow the gulf; overloaded, unranked feedback increases interpretation burden. Some state detail is limited by security or performance, so provide a user-level summary plus a diagnostic entry. Evaluation gulf complements execution gulf: one concerns reading results, the other producing actions.
Applying it
- For every state change, answer: what can users see, what does it mean to them, and can they judge goal completion?
- Express state in user objects and task language, not internal status codes.
- Give key results evidence such as object count, time, version, error detail, and undo entry.
- Use pause interviews to measure whether users can correctly state current status and completion.