Abstract: Progress bars are prevalent in modern user interfaces. Prospective paradigm, verbal estimation method and single task program are used in this research to explore the impact of change in velocity and length of duration on user duration perception of progress bars. The results are used to suggest several design considerations for intelligent progress bars that can cache progress when the operation first starts to mitigate negative progress behaviors (e.g., slow-down). Thus it makes progress bars appear faster and ultimately improves user computing experience which fully reflects the user-centered design concept of user interfaces. The Experiment 1was conducted with a 2（duration : short duration，long duration ）×4 (progress bar type : constant, decreasing, accelerating, accelerating in the beginning and decreasing towards the end) within-subject design . In order to reduce the effect of integer reaction tendency of verbal estimation, the experiment adjusted the length of duration slightly which changed among 5 adjacent durations. Short duration was about 10s and the range was respectively 9 s, 9.5 s, 10 s, 10.5 s and 11 s, while long duration was about 30s and the range was respectively 29 s, 29.5 s, 30 s, 30.5 s and 31 s. On the basis of Experiment 1, Experiment 2 chose the duration consistent of previous studies and explored the reason for the discrepency between our results and the ones of previous studies. The dependent variable was the ratio scores of duration judgments and the ratio = estimated duration / physical duration. The progress bars as experimental materials were 28 cm × 2.1 cm in size. As the trial began, participants were presented with a grey progress bar which was filled with blue gradually. They were asked to focus on the progress bar and repeat “the, the, the” quickly to avoid some duration estimation strategy (i.e., counting silently). Meanwhile the progress indicator (e.g., 21%) above the bar moved with the blue progress bar from left to right (Figure 2). As soon as the progress bar was filled with blue completely, participants input their duration estimation in a text box at the bottom and pressed “submit”. Then the text box was emptied to avoid the tendency of attainability of verbal estimation method (i.e., submitting the former result). Afterwards another trial began. Presentation order was randomized. The results are as follows: (1) users tend to underestimate long duration and overestimate short duration under on concurrent task condition ; (2) significant length effects are documented. The length of duration is in inverse proportion to the ratio scores of duration judgments. The duration judgments ration scores decreases as the duration length increases; (3) the main effect of progress bar type is significant. The scores of constant progress bars are maximum and those of accelerating progress bars are minimum.

Key words: duration perception, progress bar, human-computer interaction, verbal estimation

摘要： 进度条在人机交互界面中具有广泛应用。本研究运用单任务的预期式范式和口头估计法，探究速度变化对不同时距长度进度条的时距知觉的影响，为进度条的智能化设计提供参考。研究表明：（1）时距长度影响个体的时距估计，时距长度与时距估计比率分数成反比关系；（2）进度条的不同速度变化趋势影响进度条时距知觉，在持续时间相同的匀速、减速、加速和先加后减四种进度条类型中，加速进度条的时距知觉最短，而匀速进度条时距估计最长。

关键词: 时距知觉, 进度条, 人机交互, 口头估计法