The researchers in this article developed software and an analysis tool to collect and analyze real world pointing performance, which was used to investigate the variance in performance of 6 individuals with a range on pointing abilities. Performance pointing research is generally done in a very controlled environment (usually a lab) where the users are instructed on exactly what to do. However, this method ignores a plethora of useful data that can be gathered if gathered in a real world scenario (like users clicking links mistakenly, and going back immediately). They focused mostly on the following types of pointing errors
- Too many buttons: When a user accidently clicks both the left and the right mouse buttons together
- Accidental click: When the user clicks when they didn’t intend to.
- Double click speed: When the user does not click twice fast enough – instead of a double click, it is registered as two single clicks.
- Movement during a single click (slipping): When the user moves the mouse cursor while clicking a button.
The research study also collects data on the direction changes of the mouse pointer and the excess distance travelled (ratio of actual distance travelled by pointer vs straight line distance from mouse pointer to target). Their software tool was based on (and extends) the DART tool which is a suite of system monitoring components that run in the background to log pointing, keyboard and window events.
Overall, they were able to gather real world data that was more detailed than data gathered in a controlled environment as the users had the liberty to use the test computers when and how they wanted. They found that most of the participants did not know/were not willing to adjust the pointing settings in order to make their experience easier.
Overall, I think it was a nicely written paper. Although they talk about the reasons for studying pointing performance like measuring pointing variances between individuals of various abilities, and studying performance across various pointing devices, it was not clear what the real world use of the data could be. Can this data help them design better pointing devices? Can they use this data to recommend/design different pointing devices to people with diferent disabilities? I couldn’t really deduce from the paper, what some real world uses of the data was, and based on their analysis, I could only imagine that their data could produce better real world design than “lab-gathered” data – I just couldn’t tell exactly how it was going to accomplish that.