Insights in the eye of the beholder

Getting value from commercial eye tracking

Commercial eye tracking studies have recently grown in popularity. However, sceptics are still wondering what exactly the patterns of eye movement across an interface or website can really tell us about online user experience.

Eye tracking is not new. The first non-invasive eye tracker was developed in 1901 and researchers have been using eye tracking to investigate computer interfaces, menu design and cognitive processing since the 1980s.

However, recent developments in eye tracking technology mean that eye movement patterns are now much easier to record, easier to analyse and easier to visualise. Practitioners can offer insightful and visual eye tracking deliverables quickly and easily and can record eye movements without interfering with participants’ natural movement and behaviour using remote tracking technologies.

Heat maps visualising patterns of eye gaze for several participants can be used to demonstrate which elements of a website participants attended to and which elements were ignored. Alternatively, a users’ point of gaze can be superimposed on to a video of the website they were viewing enabling clients and analysts to understand an interface through the eyes of a user.

Eye tracking is a powerful tool that can deliver valuable insights when it is used professionally and knowledgeably. To get real value from eye tracking, it is important to decide what you need to know and how you will leverage this knowledge to make changes that will enhance customer experience, improve usability and unleash creativity.

Understanding eye movements

Our eyes move so that we can create a rich visual picture of the world that surrounds us. Light from our surrounding environment enters the eyes through the transparent cornea and then through the pupil. An area of interest in the world is then focused by the lens of the eye on to the retina, a layer of cells at the back of the eye. In the centre of the retina there is a small area, known as the fovea, which has the greatest density of light-sensitive cells. This area is able to provide us with the highest quality or acuity of visual information.

At a single instance in time, only a small part of our visual world is imaged on the high acuity fovea whilst most visual information falls on more peripheral areas of the retina which have a lower acuity. Our eyes move so that the image of the current object of interest falls on the fovea, allowing us to see what we’re interested in clearly and accurately.

Our eyes are constantly moving in a variety of different ways and patterns according to specific circumstances.

In looking at websites, the main eye movements of interest to researchers are saccades and fixations.

Saccades are rapid eye movements that occur so that a particular area of the surrounding world around us falls on to the fovea. Saccades enable us to select what we attend to visually.

Fixations occur when the eyes are relatively still, between saccades. Information about the world is picked up during fixations, when the eye extracts information from the area fixated and reflects this information onto the fovea at the back of the eye. The fovea is an area at the back of the eye responsible for our central vision, necessary for focused tasks such as reading or extracting meaning from an element in a visual display.

During fixations, our eyes are still moving, making miniature eye movements such as drifts, flicks and high frequency tremors. The effect of such small movements is to constantly shift this image minutely over the fovea so that the fovea is constantly being stimulated – if this did not happen then the image of the object in focus would fade.

Fixations and Attention

We attend to information that falls on the fovea and the parafoveal area surrounding the fovea. To create an impression of a web page, a user would typically make just a few fixations, with saccadic jumps in between to different areas of the page in order to get a gist of what the page is about. Further fixations would focus on the detail in specific areas of interest across the page. This quick scanning process enables a user to form a mental picture of the display and decide where to focus next, in order to extract more information and detail.

Large eye tracking studies on news websites have shown that users typically scan web pages before reading in depth and this approach to extracting visual information on the web has encouraged designers to create content which is easier to scan using paragraph breaks and headings.

The Eye Mind Hypothesis

Typically, we attend visually to the information that we are fixating but sometimes there is a difference between what we’re attending to and what we’re looking at. In addition, people often remember visual information that they did not fixate on, for example banner advertising on web pages. A French eye tracking study showed that although participants didn’t fixate on banners, exposure to web pages including the banners led to increased recognition of brand names used in the banners. There is a relationship between where we are fixating and what we are attending to but we also attend to peripheral information.

Eye tracking research on reading and on spoken word comprehension has shown that we can learn an enormous amount about how we process information from eye tracking data. In a carefully designed study eye tracking can reveal how we process the meaning of a sentence and to predict where we decide to look when assigned a task, before we have consciously decided how we will accomplish that task.

By combining eye tracking data with other research data, such as task completion measures, usability observations and subsequent memory measures we can gather more insight about what patterns of fixations and saccades really mean. Using a range of measures helps to determine whether a long fixation is the result of difficulty in accomplishing a task or engagement with a strong visually creative element. Combining measures helps to validate our assumptions about the eye mind hypothesis.

Understanding human cognition

Eye tracking studies can be employed to study the intricacies of human behaviour. A good example is a study that sought to understand whether experts and novices process visual information in the same way. A study investigating chess players showed that expert chess players made different patterns of fixations from novice chess players, but only when looking at configurations of chess pieces which would be found in a real chess game.. When looking at pieces that were irrelevant to a game, there was no difference in the way that players extracted visual information about the chess pieces. The experts’ memory for chess configurations is thought to have helped them to extract visual information from the chess board with fewer fixations than novices. They were more likely to make fixations between pieces, helping them to pick up interpiece relations quickly.

On the Web, patterns of fixation could also differ for experts and novices but this is yet to be tested. Research to date has shown rather different findings. A study looking at the impact of changing the position of the conventional left hand navigation to right hand navigation showed that there were only temporary familiarity effects, and that after seeing one page with right hand navigation, users found it easy to switch to focusing on the right hand area of the screen. This highlights the importance of keeping the design consistent across pages of the same site but shows that on the Web, users can easily adapt to new layouts which may have design advantages over keeping with the conventional or familiar.

Testing Design Hypotheses

Eye tracking can be employed to evaluate alternative design layouts or navigation designs for a specific website. By combining eye movement recordings with usability testing methodologies it is possible to evaluate how different layouts influence a users’ search efficiency on a site and how different elements of a site could hinder or support users in completing a set goal. An example is a study to test the positioning of a navigation menu. Users are more familiar with left hand menus; however positioning the menu on the right hand side may have significant advantages for some sites. Eye tracking studies have revealed that users soon adapt to a site which has navigation on the right and make fixations to the right rather than the left on the second page of the website.

Eye movement patterns on the Web

Sometimes, researchers use patterns in the data from large scale studies to draw more generic guideline s about online behaviour or about interaction with websites.

Large scale eye tracking studies have revealed some key findings that have shaped good practice design for the web.

  • Advertisements in the top and left of a web page attract the most attention from users;
  • Close proximity of advertisements to popular editorial content helped to ensure adverts were seen;
  • Shorter paragraphs were read more thoroughly than longer paragraphs;
  • Users may be more likely to read to the end of a long piece of text if it is well structured with short, meaningful headings and short paragraphs.

Conclusion

When using eye tracking for insights into the usability of a website, there are three main approaches you could take. Eye tracking studies can proceed top down, based either on a pre-defined theory about human cognition or alternatively based on a design hypothesis about a specific interface. Eye tracking studies can also proceed bottom up, where results are based entirely on observation of the data. The right approach to take depends on the goal of the research, the number of participants involved and the outcomes that the research is intended to shape.

Eye tracking can uncover many different kinds of result. To draw meaningful conclusions from eye tracking data requires a sound knowledge of human cognition and user behaviour. Distilling trends from the variety of scan paths and gaze replays recorded during eye tracking studies helps us to understand how users form impressions and process information on the web.