Touch screens are a popular method of interaction with information systems embedded in public kiosks and PDAs. Typical information systems are used on desktop PCs and therefore restricted to having a mouse as the selection device used to interact with the system. The purpose of this study is to investigate how effective a touch screen is in selecting typical GUI items used in information systems.
A series of tests were completed involving multi directional point and select tasks. A mouse, being the standard selection device, was also tested so the results of the touch screen could be compared and evaluated against. The GUI items tested were a button, check box, combo box
and a text box. The touch screen used was a 17" Magic Touch USB overlay. During the test, data relating to movement time, error rate and throughput were captured.
The results showed that the touch screen overlay was not suitable in terms of selecting small targets with a size of 8mm or less. There was a significant difference between the touch screen and the mouse in terms of movement time and error rate. The touch screen overlay was 15% slower than the mouse and had a 61% error rate compared to the error rate of 2.7% for the mouse. There was no significant difference in throughput between a touch screen overlay and mouse. Both selection devices had throughput of 1.2 bps.
User satisfaction relating to each device was evaluated using responses to a device assessment questionnaire, completed by each participant at the conclusion of the test. The mouse was rated easier to use and easier to make accurate selections with. Both results were statistically
significant (p < 0.05). The question regarding arm, wrist and finger fatigue was also statistically significant with the touch screen rated worse than the mouse.
The report concludes with some practical implications of the study; namely that a touch screen overlay used only with a finger is not a practical selection device to use with typical information systems.
