Abstract
Introduction
Experiment
Results
Discussion
Conclusions
Acknowledgements
References
Appendices
Credits
Feedback
Back To Main
A Tail of Two Mice
Conclusions
Average performance times show two trends consistent in both mouse treatments. The first trend was the improvement of performance times in the second trials. The roller mouse group's time decreased by 18%, and the non-roller group's time decreased by 9%. The other trend was the drastic increase in performance times in the third trials. The roller mouse group's mean time increase by a staggering 33%, and the non-roller mouse group showed a 30% increase in completion time. In other words, the time to complete the task first dropped moderately, then raised considerably. What is truly interesting is that the average times for the third trials were actually more than the average times for the first trials! (Roller-mouse group up by 9%, non-roller mouse group up by 18%). Therefore, we are seeing both sides of the spectrum: decreases and increases in the time to complete the tasks. This information has lead us to redefine our theory and make suggestions for future research.Redefining the theory
Our previous hypothesis states, "The times for the successive tasks completed with the 'scroll wheel' mouse will improve, while the times for the successive tasks completed with the conventional mouse will remain about the same." This must be changed, for it is only partially correct. Instead, a critical threshold of productivity must be understood before the duration of an experiment is set. In other words, how long can users perform before they become tired and/or frustrated? Judging by our experiment, that threshold of productivity is approximately 6-7 minutes of continuous searching time. After that, the users' eyes become fatigued and they report higher levels of frustration. This certainly attributes to their decreased performance.Suggestions for future researchers
On that note, future researchers must design experiments while keeping in mind the balance between productivity and diminishing returns. If not, then the results are misleading, because the higher times of completion may not be attributed to the mouse itself. Rather, other factors (like frustration) are correlated to poorer performance. In addition, issues surrounding the scrolling lists must be considered. The lists must be totally random during all trials to prevent biases between the lists. This design is more realistic, because conducting Web-based research may include unpredictable events. That way, our research will be of use to both manufacturers and the average user.
Department of Computer Science: Direct questions and comments to the student editorial team
