The Use of Eye-Trackers in Research
They say you can tell a lot about a person based on their eyes, that eyes are windows to the soul. We can tell someone’s emotional state based on the expression in their eyes. We can tell if someone is lying to us based on abnormal eye movements. We can tell if someone is romantically interested in us based on pupil dilation. To many researchers in industry and academia, eyes are mirrors of the mind – a reflection of cognitive processes within, and these mirrors are eye-trackers.
Eye-trackers are devices that track a user’s eye positions and movements. They use a tiny, unnoticeable infrared light and its reflection on the user’s eyes in order to infer the location or direction of the user’s gaze.
The use of eye-trackers has become more and more popular. There are two types of eye-trackers available now:
1. Wearable devices (Eye-tracking Glasses)
2. Stationary devices (Screen-based Eye Tracker) that are mounted to a computer screen
For most research purposes, screen-based eye-trackers would suffice. They work well with two-dimensional stimuli, such as images, videos, websites, games, that are presented on a screen of a computer monitor, phone, or tablet. They could also be used with books and magazines, provided that the user is not moving a lot. For interactions with three-dimensional objects and navigation of environments, eye-tracking glasses are recommended because they move with the user.
Eye-trackers collect a lot of data on users’ eye positions and movements while performing tasks. The most important ones are described below:
Key Eye-Tracker Terms/Metrics
1. Fixations or where the user is looking at. The most common metrics people talk about are metrics related to fixations. For example, number of fixations on a screen, the duration of each of those fixations, and revisits, or locations of repeated fixations.
2. Areas of Interest (AOIs). Just like neuroscientists have regions of interest, where they focus their data analysis, researchers who use eye-trackers have areas of interest. These are regions on the screen that they are summarizing metrics from. Depending on the research question, different stimuli are clustered together to make an area of interest. For example, if you’re interested in how people observe body language and make social judgments, you may want to cluster all eye-tracker metrics regarding a person’s face separate from a person’s body.
3. Time to first fixation, or the amount of time until the user fixates within an area of interest. This metric tells us how information is being prioritized when viewing the screen and helps researchers gauge the salience of different objects on the screen.
Beyond looking at the sheer numbers generated by these metrics, researchers also make use of data visualizations to quickly assess user behavior and to communicate their findings. Saccades can be summarized using fixation sequences or a mapping of a user’s fixations in the order that they happened on the image that the user was viewing. Another useful data visualization is heat maps. They show the distribution of fixations across the screen with areas more attended to displayed in red.
Eye-trackers are exceptionally useful to researchers in industry and academia.
Industry. They are used in industry for consumer research, most often for marketing and web design. Researchers use eye-tracker metrics to assess the effectiveness of certain marketing strategies as well as the ease of user experience for websites or software.
In terms of marketing, researchers may be interested in the time to first fixation of the brand name on a product and perhaps the duration, as marketers want the consumer to remember the product. Eye tracker metrics also helps them identify any distracting features that steer the consumer away from the important message or the brand label.
Academia. Eye-trackers allows researchers to assess existing theories of mental representations in a more “physical” way. For example, researchers, who study numerical cognition, have long agreed that people represent numbers in an analog fashion, much like a mental number line. This is an inference made based on participants’ error rate and response times when comparing magnitudes of numbers. Eye-tracking data has provided more convincing evidence for this theory. For example, researchers(1) asked participants to generate “random” numbers, one after another, onto a computer screen. Based on whether participant’s gaze shifts left or right, we can predict whether they will generate a smaller or bigger number than the one they already generated. It appears that participants are referencing a mental number line by finding numbers that are less than the current number on the left and finding numbers that are greater than the current number on the right.
Eye-trackers contribute a lot to research in the domains that I have discussed. Most importantly, it is objective and gives us more details about the user than the user can self-report to us.
(1) Ruiz Fernandez, S., Rahona, J.J., Hervas, G., Vazquez, C., & Ulrich, R. (2011). Number magnitude determines gaze direction: Spatial-numerical association in a free-choice task. Cortex, 47(5), 617 – 620.