Читать книгу Eye Tracking the User Experience - Aga Bojko - Страница 36

There are two main types of eye trackers used in UX research: wearable (see Figure 3.1) and remote (see Figure 3.2). The main difference between the two is the location of the equipment during tracking. Wearable eye trackers are worn on a participant’s head, while remote eye trackers are contact-free and placed in a fixed location in front of a participant. Table 3.1 compares remote and wearable eye trackers based on their uses, advantages, and limitations.

FIGURE 3.1 Examples of wearable eye trackers.

FIGURE 3.2 Examples of remote eye trackers.

TABLE 3.1 A COMPARISON OF TWO MAIN TYPES OF EYE TRACKERS

	Remote Eye Trackers	Wearable Eye Trackers
Setup	Placed in a fixed location in front of a participant (e.g., on a desk or a car dashboard).	Worn on a participant’s head (e.g., as a pair of glasses, attached to a hat, or on a headband).
Application	Used in studies during which participants can sit or stand in one place and the stimulus is presented on a stationary surface (e.g., research using on-screen stimuli such as websites, images, and video).	Used in studies that require participants to move around and interact with physical objects or people (e.g., wayfinding, shopping, and out-of-the-box research).
Obtrusiveness	Less obtrusive than wearable systems—participants can easily forget about the eye tracker during the study.	More obtrusive because the system has to be worn on the head and is usually partially visible to the participant (even though people learn to ignore it after a while). Also, headgear may be undesirable due to hygiene or hairstyle concerns.
Freedom of Movement	Participant must be positioned in front of the eye tracker and only limited head movement is acceptable. There also has to be a clear path between the participant’s eyes and the eye tracker (e.g., a participant cannot hold anything in front of his face).	Participant can move around freely and manipulate objects. However, most wearable eye trackers work best for objects that are at the same distance as the calibration, and are less accurate for objects that are closer or farther. This is called a parallax error and only some systems can correct for it.
Ease of Analysis	Because the recorded scene doesn’t change with head movement, the same gaze location in the recorded frames indicates the same object in the world. Data analysis is typically easier and faster because more automated data mapping and aggregation are possible. However, that’s only the case for static stimuli. If the content is dynamic (e.g., a video or a website with a lot of overlays), the lack of a stable coordinate system can make the analysis more manual and time-consuming.	The recorded scene keeps changing due to head movement. As a result, the same gaze location in the recorded frames can indicate something different in the world: Because of this dissociation, data analysis is typically more manual and time-consuming. This problem can be at least partially solved by placing infrared markers in the environment, using edge/object detection, or adding a head tracker.