Goals for User-Interface Design

The following five goals lead to qualitative aspects of user-interfaces that can be measured through user-testing:
  1. Time to learn:
    The user-interface should be easy to learn. The time it takes an "average" user to learn how to perform a given task using the user-interface can be measured.

  2. Speed of performance:
    A skilled user should be able to perform a given task in as little time as possible. This aspect can be partly evaluated without users, by enumerating the steps needed to perform a task (through the user-interface) and adding the times for each step. Such a preliminary evaluation can be confirmed by tests with users.

  3. Rate of errors by users:
    Users should make few or no errors when using the interface to perform a given task. Although errors contribute to lower performance, they should be measured separately. Frequent or repeated errors indicate that the user-interface creates confusion about what the effect of certain actions will be.

  4. Retention over time:
    Users should be able to easily remember how the interface works and how they use it to perform a given task. The performance and error rate of users can be measured after not having used the system for different time intervals, like an hour, a day, a week...

  5. Subjective satisfaction:
    The user should like using the interface. They should feel that they master the system. Subjective satisfaction can be easily measured by means of user surveys. It is often less important how well the user masters the system than how well the user thinks she masters the system.
It may not be possible to score well on all five goals, and they may be in conflict with each other. If a more lengthy learning period is considered acceptable, task-performance may be improved through a more complex interface. In such a case retention over time may be less.
Subjective satisfaction is often very important because users may eventually end up performing better using an interface they feel comfortable with.