Eyetracking
Usability tests often use measures designed to evaluate:
- Effectiveness
- Efficiency
- Satisfaction
While these measures are especially effective at indicating what occurs in the context of using a device or design, they are not always effective measures for determining Why users performed in a particular manner. This is especially important in the context of more complex interfaces, such as games, where the reasons a user may fail or take more time to perform a particular task may not be apparent through usability results alone.
So Why Eye Tracking?
In cases such as evaluating games, eye tracking may prove to be a valuable asset in determining what the user is doing during a game session. Eye tracking may be used to observe:
- User Strategies
- Interface Design Issues
- Differences in Expert-Novice Performance
- Attentional Behavior Leading to Errors
- Challenge (Workload)
Our group is working on several approaches to evaluating gameplay and user strategy using eye tracking. Below is an example of using eye tracking with Age of Empires 3 to identify a usability issue.
For example, this particular game uses an icon in the upper left corner to identify when settlers (units responsible for collecting resources and building structures) become available for use. While running several novices on the game, we noticed that some did not use the icon as they performed various in-game tasks. Our cognitive walkthroughs confirmed that the user did not verbally reference using the icon. When questioned, however, the user responded that they were aware of the icon and visually referenced it during the game.
Where would this leave your average usability researcher? With data inferring one thing and an anecdote pointing in a different direction, and no solid information to determine which is more accurate. Rather than wager a guess, we can look at the user’s eye movement data:
Imagine each red dot on the image accounts for a single fixation point. The locations of the fixations can be used as an indication of the location of visual attention. Note that throughout the session, the user does not actually fixate on the icon, and only once comes in the vicinity of the icon. This makes a strong argument against the user actually referencing the visual icon.
Given this information, we were able to discern that the user did not actually fixate on the on the icon, and provided recommendations for improved icon design for drawing the user’s attention.
Our current work looks to expand the use of eye tracking in games to include multiple types of games, as well a template of scenarios where eye tracking could be expected to provide additional insight to usability results.