6.5.9 Self-descriptiveness measures

What this heading covers

A self-descriptive interface lets a user act without documentation. The measure is about comprehension, so it needs a user and a question: what does this screen tell you, what happens if you press this, what is this value. The population of screens should include the ones where a mistake is costly. The measure is cheap to collect with a comprehension test and expensive to fake, which makes it a good candidate for a regular check.

Record position: family 16 of 40 in clause 6, under 6.5 Interaction capability, which carries 8 families. It is one of the 40 families a programme has to map its measures to, and its elements are listed below. The heading is 6.5.9 in the public contents; the text behind it is not public.

How a measure for this family is built

Take the number of comprehension questions answered correctly as the numerator over the questions asked, across the screens and users in the study. Report the screens that failed, because the average hides the one dangerous screen. A second measure is the time to a correct interpretation.

What to put in place before publication

1. Choose screens where a misunderstanding would matter.
2. Ask what will happen, not only what is shown.

Evidence to keep

Two questions to ask

• Which screens were tested, and were they the ones where a mistake is costly?
• Did the questions ask what will happen, or only what is shown?

Where teams usually fall short

• A preference survey used as evidence that the interface explains itself.

What the number cannot tell you

It cannot tell you whether the interface is pleasant or efficient. It tells you whether it was understood.

What happens next

This family completes interaction capability. The reliability families are next, and they change the evidence from users to operations.