The Effect of Zooming Speed in a Zoomable User Interface

Authors:

Huo Guo-hguo@cs.umd.edu
Vivian Zhang-suje@wam.umd.edu
Jing Wu-jingwu@cs.umd.edu

Abstract

This experiment was conducted to measure the effect of the zooming speed in a zoomable interface. The independent variable was the zooming speed. The dependent variables were performance time, retention accuracy and subjective preference.

We hypothesized that there would exist a zooming speed at which the performance time is shortest. We also hypothesized that retention accuracy would improve as zooming speed decreased. The third hypothesis was that subjects would prefer the zooming speed at which the performance time was shortest.

Thirty-nine subjects participated the experiment and five were dropped from analysis due to interruptions in the course of experiment. The subjects used a zoomable interface system to finish twenty-four searching tasks. They answered six retention test questions and one subjective preference survey question.

The outcome of this study confirms our belief that there exists a zooming speed at which the user performance is best. The experiment shows that shortest performance time can be achieved at the zooming speed of 8 factor/second when taking the search tasks in our zoomable user interface. No statistically significant difference of retention accuracy was observed. Subjects were most dissatisfied with the slow zooming speed.

Based on the experiment result, we suggest researchers and practioners consider using the zooming speed 8 scaling factor/second.

Department of Computer Science: Direct questions and comments to the student editorial team