When conducting eye tracking analytics, the first step is understanding what metrics are important to you, which is influenced by the goal and questions you want to answer. On this page, we present some common, high-value metrics you should be considering in your data collection efforts.
You can learn more about how our eyes work, including different eye movements, and how these are related to the raw signals of the eye tracker in our eye behavior section.
Table of Contents
Often, fixations on areas or objects of interest are the most important metric to help you answer your questions. They can teach you what grabbed the attention of the user first, how interested the user was in a specific object or if something was difficult to understand or process. A fixation can be calculated as a set of gaze points generated by the eye tracker in close proximity in time and space.
Here are some valuable fixation-derived metrics that can be used to gain insights:
Fixation Duration: To understand how users perceive objects and areas of interest, fixation duration can be used. Higher fixation durations can signal a higher level of interest, whereas shorter fixation durations might mean other objects or areas of interest are more attention-grabbing or interesting. It’s hard to determine emotional reaction based solely on this metric though, and requires corroborative metrics and data. This metric is calculated by the total time fixated on an object or area of interest.
Fixation Count: A high number of fixations, meaning the user repeatedly revisits the same object or area could mean a high level of interest. However, it could also indicate difficulty with comprehension. While this metric cannot tell you how users feel when looking at something, it provides data as to what should be examined further.
Average Fixation Duration: Through a simple combination of duration and count, you can see the combined outcome of both metrics. This can allow for quick insights into perception but can also be used a baseline for both fixation duration and fixation count. Having a higher average fixation duration for an object relative to others is probably preferred if it is something you wish for users to notice and take in, for example an important message or a new product.
Time to First Fixation: This metric is useful when analyzing how quickly something captures your attention. For example, if you want to want to know the optimal placement for an information sign in a store, you could compare different sign designs and placements to see which one grabs users’ attention the fastest. This metric is calculated based on how long it took for participants to notice the object or area of interest from the start of the scenario.
First Fixation Duration: This can be used in conjunction with time to first fixation to help determine the first impressions of objects or areas of interest. For example, if the time to first fixation is low and the first fixation duration is high, the area or object is likely perceived as very attention-grabbing for the user.
In the below example, Time to First Fixation, Fixation Duration and Fixation Count is being recorded for shoes.
Pupil Size: Changes in pupil size are useful as an indicator that the user is having an emotional or cognitive reaction to an object of interest or event. However, pupil size can also change for other reasons (primarily lighting changes), so this need to be accounted for. Using this metric in conjunction with other data points, like asking participants after the study about moments where the pupil size changes occurred, may lead to interesting insights into users’ internal thoughts and feelings.
Distance to Object: Since the clarity of details can vary between distances, it can be useful to include this in your metrics. A person spending a lot of time looking at a page of a book near vs. far is a big difference in experience and should be reflected in the data.
Viewing Angle: Objects can be perceived differently from different angles. For example, a person may see a billboard from the side at a shallow angle, and not be able to read the text. This would give the user a very different perception of the billboard, compared to seeing it from the front and clearly reading the text.
Gaze Angle: Depending if a fixation occurs in the periphery of your vision or in the center of the field of view can play a part in why we are paying attention. Since we are better at detecting changes of speed in our periphery and we tend to align our head with our gaze direction in times of focus, quick fixations of objects in the periphery might be seen as distractors whereas objects in our center view may be more indicative of conscious focus.
The video below shows how pupil size changes according to lighting changes in the scene. When lighting levels are controlled for, changes in pupil size reveal emotional or cognitive reactions.
Combining with other Metrics
Eye tracking analysis is best done in combination with other measurements in order to get a clearer picture of user behavior and the context of the eye movements.
Interaction Metrics: Recording metrics related to interactions or events may be valuable to look at in conjunction with eye tracking data. For example, if a user starts interacting with an object immediately after seeing it, they may have found it more interesting than if they have looked at it several times in the past before starting to interact with it.
Other Biometrics: Other biometrics like GSR, heart rate, EEG etc. can be used together with eye tracking to find better or more results. One example is to use heart rate and EEG together with changes in pupil size to more accurately be able to detect emotional and cognitive response from stimuli.