Mouse tracking

We wanted to find some interesting analogue data which could be collected and analysed quickly, and for this analysis to produce reasonably useful real world output. Well, we didn't have to look much further than the humble computer input device, the mouse (or in this case, a trackpad. All recordable by products like IOGraph). The goal of this analysis is to suggest improvements in the application design, screen layout, workflow, and of course the habits of the end user themselves. This is done by studying how the user interacts with the computer while performing ordinary, everyday tasks.
Here are the broad constraints of the data gathering exercise.
  • Approx 9 days of casual activity, from a volunteer using an Apple 13 inch laptop, using the inbuilt "trackpad" device
  • The user is experienced with computers
  • Computer was left on, never rebooted
  • The majority of usage is Internet browsing/twitter etc
  • Small amount of document creation, email, word documents, spreadsheets
  • Small amount of photo browsing
  • Main apps are Web browser, email client, Twitter client, and an IRC chat client
  • itunes running
This animation shows cumulative movements over time.

Some interesting points can be made. There are some obvious hotspots where the mouse was drawn to:
  • The top right is the screen lock "hot corner" where the user moves the mouse to lock the screen while away from the keyboard
  • There is a horizontal band towards the bottom right. This was due to the Tweetdeck framework not supporting horizontal two finger scrolling
  • The grouping at the bottom is the application dock, where user went to swap between applications
  • The clusters at the top and the middle are mostly email checking, and pushing "favourite" buttons inside the web browser (eg news)
Of course this data is very specific to the surfing habits of the user. The point is that efficiencies can be increased by changing some habits, improving workflow and application layout, and by introducing some useful features of the system itself (eg the multi finger gestures offered on the Mac). Examples include example keyboard shortcuts, and numerous trackpad gestures. The layout of the applications could also be changed to decrease the travel distance of mouse movements.

This concept can be stretched further to study highly repetitive task-orientated computing environments. Examples include Call centres and data entry environments where slight improvements in application design and usage patterns can produce significant gains in production, not to mention support higher levels of user engagement.
Copyright dataviz Australia 2010