Duration estimates and users' preferences in human-computer interaction

Two experiments assessed the effect of various static and dynamic computer 'wait' message displays on (1) subjective estimates of the duration of intervals during which a subject had to wait for the computer's response, and (2) subjective preferences among the different displays. All the static displays led to identical duration estimates. For dynamic displays a direct relation between the rate of changes and the estimate was found. Faster rates led in most cases to increased estimates of duration, and slow-changing graphic displays appeared to have the shortest duration. Subjects preferred epigrams to all other displays, and cumulative graphic displays to a blinking or static WAIT. The rate of change did not affect preference ratings for the graphic displays, whereas slower blink rates were preferred for the WAIT. The results demonstrate that the findings and models from earlier basic research on time estimation are relevant for predicting the subjective assessment of wait-periods. However, users' satisfaction with the displays depends not only on the apparent duration of the wait period, but also on other variables. These should be considered when choosing the display that is shown while users wait for the system to complete a task.