When you want to know what the main attractors on your homepage are or how your users are moving through your site, you should consider eye tracking. Eye tracking is a fun method in a design researcher’s toolbox. While users interact with a website, researchers record where their eyes go along with other data such as where their mouse goes. In particular, we capture the path their eyes take through the site and measure how long they linger on certain areas of the screen. And the best part is the fascinating heat maps and tracking maps that we get at the end. Like anything else, though, it’s not a one-size-fits-all method, so let’s take a moment to break down what eye tracking is really good for.
When Should We Use Eye Tracking?
Eye tracking is great for:
- General usability studies to understand efficiency and effectiveness of a site
- Studies focused on usability of complex interfaces (Why aren’t our users completing their core tasks?)
- Studies related to branding and content strategy (What’s catching a user’s attention? Are they even reading anything on our site?)
- Making comparisons, such as A/B testing two versions of your site, comparing your site to a competitor’s site, or considering how different target user groups interact with your system
What Is Eye Tracking?
When we conduct a usability test, we’re asking research questions about how well users move through a site, how effectively they complete key tasks, and how satisfied they are with their experience. We’re also asking questions about how quickly users learn a site and recover from their mistakes (i.e. ‘learnability’).
Eye tracking helps us address these questions by showing significant areas of a site that users miss, such as a critical calls to action or key steps in a workflow. As users progress through tasks, we can use eye tracking to determine if they are able to learn the layout of the site quickly to accomplish core tasks. In essence, it helps researchers understand the possible causes for user performance by providing data about the location and duration of visual fixations and attention.
How Does It Work?
Eye tracking technology has evolved over the last decade. The equipment is now small and portable, so that we can study the use of websites even on mobile devices. Before the study begins, we connect gaze tracking technology to a laptop, tablet, or other device and use some behind-the-scenes software together with the device’s camera to capture all the data. The user participates in a brief eye calibration step and then it is pretty much business as usual – the eye tracking technology captures data unobtrusively, A user goes through some tasks, gives us some ratings, and participates in a post-study interview.
What Do We Do With The Data?
Eye tracking data on its own isn’t terribly meaningful. It needs to be combined with other types of data such as interview or survey data. Like most quantitative methods, it’s good at telling us the ‘what’, but it can’t tell us the ‘why’. So we often use multiple methods that balance quantitative and qualitative data to give us the fullest picture. The data output includes video playback of a user’s visual path overlaid onto the web interface as well as heat maps and path maps to identify patterns across users.
But Wait, There’s More!
You don’t have to wait until you’ve completely built something to use eye tracking. You can also use eye tracking for exploratory research studies to think about innovative designs and technologies. Let me share an example from one of my favorite eye tracking studies that I conducted.
A few years ago, I ran a study with some colleagues from the University of Maryland’s HCI Lab. We wanted to improve the design of a system for social tagging of art works. Projects like the Steve Tagger were gaining traction in art museums to augment the metadata associated with the images in their digital collections. The goal of social tagging projects like the Steve Tagger are to improve search and discovery of new art by visitors to a museum’s digital collections using words that are more natural to people. For example, we may not remember the name of a painting we saw once, but we may vividly remember a little girl in a red raincoat holding an umbrella in that painting.
For our study, we were curious what people were looking at when they decided to tag an image, especially for nondescript abstract art that can look more like blobs of paint than anything else (no offense to the abstract artists out there). Eye tracking with follow up interviews was the ideal approach for addressing these exploratory research questions. We conducted an experimental lab-based study where we had individuals look at images of art and tag them. We were able to use the eye tracking data to see what they looked at and then what they tagged. For more traditional images with people, places, and things in them, participants tagged iconic figures in images first (e.g. the Virgin Mary) and then moved to objects before looking at the painting as a whole. For those blobby abstract art images, participants tagged colors first and then moved to emotions. The follow up interviews also helped us understand why some participants tagged some images less (e.g. “There is way too much going on for me to figure out what to tag”).
These results gave us unique insights into people’s tagging behaviors of images that had implications for the design of tag recommender systems and tag collection systems. You can find out more about it by checking out the paper here.
Are You Ready to Get Started?
Eye tracking is an exciting method that can add value to research studies. Our use of it in the field of user-centered design is still just scratching the surface in terms of the types of questions we can use it to help answer. Whether you’re considering an exploratory research and design strategy project or a classic usability study, eye tracking could help you dig deeper into your users’ behaviors. Our research team at IC is happy to talk with you more about eye tracking or other approaches to see what the best cocktail of methods is to answer your questions.
Golbeck, J., Koepfler, J. and Emmerling, B. (2011), An experimental study of social tagging behavior and image content. Journal of the Association for Information Science and Technology, 62: 1750–1760. doi: 10.1002/asi.21522