Eye tracking metrics are a valuable set of tools that can be used to uncover insights regarding participant behavior and mindset during various situations. Tobii Pro Studio and Tobii Pro Lab offer a wide selection of metrics covering fixation, visit, and mouse-click related questions. An important first step when choosing metrics is to decide which measures relate to the research questions of interest. This will also provide some indication as to the approach to analysis and sample size requirements.
Pro Studio and Pro Lab offer 12 fixation-related metrics that the researcher can choose from. A fixation is a period of time when the focus of the participant’s gaze is relatively still on an area and taking in information about that which is looked at. Fixations range in duration from 60 ms during reading to several hundred milliseconds when examining a photograph or image. Some of the most commonly used fixation-related metrics are time to first fixation, fixation duration, and fixation count. The researcher can use fixation metrics to learn what grabbed the attention of the participant first, how interested the participant was in a certain image, or if a sentence was difficult to understand or process.
Example: The study involves a web company testing a new logo design. The goal is to decrease the time it takes visitors to first see the company’s brand logo. The company is attempting to capture attention from visitors more quickly via a fresh logo. In this case, the best metric to use would be time to first fixation. The company would make recordings with a viewer seeing the existing logo, and then other participant recordings on a variation of the page displaying the new logo. Using this fixation metric, the company could find out exactly when the visitor first noticed the logo and determine if the redesign resulted in a measurable improvement in visibility and initial attention capture.
Presentation Tip: “This time to first fixation output shows the time that passed before the visitor first noticed the logo on the webpage. As you can see, the viewer noticed the new logo significantly quicker than the old logo. The new logo was actually the first thing that attracted attention once the page loaded which suggests instant visibility and appeal. The viewer reported that the updated design was more pleasing and did increase product retention. The viewer also noted that the new logo seemed to freshen the brand when compared to the old logo design.”
The next group of metrics deals with visits (or dwell time). A visit is defined as the period of time when a participant first focuses on a region until the person looks away from that region. A visit consists of at least one fixation but could include dozens depending on the size and content in the region. The complexity, color scheme, and tone of the region can all influence the visit metric. Visit duration, visit count and total visit duration are frequently used metrics in Studio and can be informative when examining participant interest or ease of understanding.
Example: A study where a company changed their slogan and wanted to see how customers reacted to the new wording. Visit count and visit duration would be useful metrics since the study is examining user attention and cognition. Repeated visits to the slogan text could signal difficulty with comprehension.
Presentation Tip: “These two visit-related metrics show the differences in looking behavior based on the version of the slogan used. In terms of visit duration, the viewer looking at the new slogan spent less time on the wording and more time looking at the actual product. This could signal quicker comprehension of the slogan and approval of the language. Because the viewer looking at the old slogan had a longer visit duration and spent less time looking at the actual product, confusion over wording may have been an issue. Considering visit count, the viewer looking at the old slogan had more visits between the slogan and product which could signal difficulty in understanding the relation with the product.”
Mouse-click related metrics are another useful measure. These metrics examine details relating to when the participant actually clicks on an area. Percentage clicked, time to first mouse click, and mouse click count are used to gain insight on attention and interest cues from the participant. The metric “time from first fixation to next mouse click” is especially useful in evaluating whether the target, whether button or text, is understood. A longer interval may be suggestive that the participant wasn’t sure what they were looking for was what they needed. For example, a button that doesn’t look like a button might not invite the click until the viewer confirms that there really isn’t anything else on the page that could do what they need.
Studio also provides a more detailed metrics reporting tool in the Data Export tab. The comprehensive data output from the Data Export function can be used in other analysis/statistics software based on individual needs. In addition to the looking and click metrics we already discussed, pupil size is also available from this tool. By looking at changes in pupil size, researchers can examine possible emotional reaction cues. It is important to remember that the pupil may change in response to multiple factors and not just those test features we might be interested in. Along with interest; lighting, stress and fatigue can all cause changes in pupil diameter. Because workload and stress can impact pupil size, the researcher must ensure that both images being compared require an equal mental workload from the participant if emotional response is the effect of interest.
Example: A study where a car brand is changing their external design and wants to see if consumer interest or desire is increased. Does it make the prospective car buyer crave the car? Since an emotional response is being measured, pupil diameter can be a useful metric for this scenario. The researcher would show two versions of the car design, the current and the proposed design. Changes in the pupil diameter are measured while tracking the participant’s gaze over the car and are then output via the Data Export function. From measurable changes in the pupil diameter, inferences about participant desire can be made.
Simply put, the metrics contained in Studio offer a way to measure responses in visual behavior and to compare changes or differences in these responses to allow one to make inferences about visibility, intelligibility, interest, or desirability. Metrics are a great way to triangulate the findings made by researchers and add support to behavioral insights.