Eye tracking

No. You have to draw an explicit distinction between resting your gaze on an element and taking the information in. Identifying a gaze point (or gaze plot) on an element does not let you conclude that the user has read the information. Perception and attention are two different cognitive processes.

The heatmap is a very widespread example of visualisation. By superimposing an image with a participant's gaze tracking, it shows the way those areas were looked at by the user. You often find heatmaps with a colour gradient (hot = a lot of time spent; cold = little or no time spent). You can also draw heatmaps that aggregate the data of several participants.

There are several systems, more or less elaborate. You can do basic tracking with a webcam and an image processing algorithm. There are standalone devices you can plug into a computer to follow gaze position on a screen. There are wearable devices in the form of a pair of glasses, which offer an interesting trade-off between measurement quality and complexity (but at a price that is still high). Finally, there are multi-camera devices you can install in a multi-screen setup (e.g. cockpit, control room, simulator, etc.). They are far more precise but require “laboratory” conditions and call for more advanced expertise.

While the technical principle is comparable, there are different types of eye tracker (built into a screen or directly into a pair of glasses). In lab conditions, you can allow yourself devices that are more invasive but very precise. Above all you control the other factors that influence the measurement (e.g. the brightness of the room). That level of rigour makes eye tracking more complex to use but lets you go further in the analyses. In UX research, people tend to use wearable eye trackers or remote eye trackers. Precision is lower, but it is enough to determine which elements are correctly perceived and which are not.

No. First because you do not pull out an eye tracker the way you whip out a guerrilla test. And then, eye tracking only gives information about the elements the user's gaze has passed over. That is useful, but it is often too restrictive for an analysis. So you can run user testing with an eye tracking setup, to have other information on top of what a test collects.

In absolute terms, eye tracking is valuable for all interfaces. Depending on the interface, you will choose one technique rather than another. For everything that happens on a computer screen, infrared bars are very effective. For touchscreen kiosks or tangible interfaces (e.g. signage), a glasses-type eye tracker is preferred. On mobile, the glasses version can be used too, though you will lose some precision.