In my field, I’m particular concerned with testing usability for online services, such as websites and electronic databases. But how can we measure how usable something is? Sure, there are a variety of testing procedures, such as Nielsen’s heuristics evaluation. This type of testing is important, however, it’s subjective and can’t be empirically measured.
There is one way to scientifically measure exactly how someone’s eye moves across a webpage, to see exactly what the person is looking at while performing web tasks. That would be through eye tracking studies.
Eye Tracking is a technology that uses a camera to measure the movements of a person’s pupil as they perform particular tasks. According to an article by Lynne Cooke in Technical Communication, there are two ways to track eye movement. The first is from a camera mounted to someone’s head, looking like some mix between an orthodontic device and a reality TV camera. This takes into account head movement as well as eye movement. The second way to track eye movement is by a device that is placed remotely. In this case, however, users must be careful to keep their heads very still.
According to the same article, the head-mounted camera method has been used to study the eye movements of pilots as they go about their tasks in the cockpit (456). These results enabled the redesign of cockpits to improve usability and ensure the reduction of human error. This is an area where usability is so important, yet is often not thought about.
What kind of information is obtained using eye-tracking studies? Software collects the data which can then be organized. According to Cooke, the data is broken down into fixations and saccades “that can be visually represented onscreen (457).”
- Fixations – how long someone looks at something
- Saccades – the eye moving from fixation to fixation
- Scanpaths – the sequential organization of the above to. The line that the eye travels.
When this kind of study is done while a person is navigating a webpage, researchers can see “how people visually progress through a page(457).” It’s implied that whatever a person is gazing at is directly linked to what they are thinking. This gives greater insight to the thought process in webpage navigation than through verbalization studies (458). In verbalization studies – where a person narrates their thought process as they work – it is not always evident where their gaze is. Also, a person can click faster than they can verbalize. Eye-tracking studies give empirical, verifiable data behind the use.
How can the above be used in web and interface design? Well, it allows designers to understand how people read online, how their eye tracks across the screen. I find Frank Spiller’s comment “The reality is that eye-tracking, while valuable, doesn't make usability testing any more powerful. It's what you do with the observations and the usability test data that counts” most apt.
Spiller points out in his blog entry that eye-tracking research must be made meaningful. He cautions users to focus on what they are trying to learn and to match what users are actually doing with the eye-tracking reports. He also states, most wisely, in response to the fact that ads in the top and left portions of a homepage received the most eye fixations that “I wouldn't recommend putting ads there. Just because they receive eye fixations doesn't mean they put a smile on the user's face.”
What this means is that eye-tracking studies alone should not exist in a vacuum. Designers must not only be aware of what users are looking at when they view their webpages, but how they react. Also, audience is very important, as are changing web trends. For example, in Cooke’s article, she states that eye-tracking studies corroborated Nielson’s supposition that people prefer text over graphics as entry points into web sites.
This study was done in 2000. Seven years later we have the web 2.0 generation, who view websites very differently than those who were studied nearly a decade ago. So the results of this study might not necessarily be applicable since users expectations of what the web is has changed.
Practical Example
In Cooke’s article, one of the findings indicate that eye movements follow a “Z” pattern. People started at the upper left corner, across to the right, “then scanned the page in small ‘z’ patterns progressing down the page…scanning…continued up the right column of the page…(460)” So the very last thing a user would look at would be the right hand column.
I’d like to relate this finding to a particular issue we were having at our library. Lexis-Nexis, an electronic database that provides access to legal materials, recently redesigned its interface. Although the page is much improved aesthetically, based on the number of issues we are finding, the usability of the new design was not well thought out.
For example, the screencap above shows the main search function of the legal resource page. A normal user scanning the page and intending to do a search, will see the sources section I’ve circled in red and stop there, thinking the drop down menu will cover all the possible sources. However, since the right hand column is the very last thing users see, according to eye-tracking studies, users will miss that in order to change the source to either case law or statutes, they need to click on the links on the very right hand side.
Also, if the user wants a specific state code, they cannot even access it from the main search screen. They need to click the sources tab and pinpoint the particular state and then return to the search screen.
I have had numerous students call asking for help in navigating the new interface. We even had a professor who was a lawyer call, unable to find the way to limit her search to NJ statutes. It took two librarians (one of whom is also a lawyer) to figure out how to change the source. And if I hadn’t attended online training on the new interface, I’m not sure I would have figured out where to go for specific state statutes.
The design fails on some of the usability principles indicated here at usability.gov:
Ease of Learning – New users to Lexis-Nexis find it difficult to quickly determine how to get the information they need.
Efficiency of use – Experienced users of Lexis-Nexis will be able to determine the steps needed, but the steps are not efficient.
Error frequency and severity – based on the number of calls and complaints we’ve received, the frequency of errors is quite high.
Subjective satisfaction – Users won’t like using the resource if they need to jump through hoops to figure out how it works.
Perhaps if the designers of the new interface had used eye-tracking studies, or even a heuristic evaluation, they might have reconsidered the placement of the Legal Searches section. If they had, it would have saved those of us at the library a lot of headaches.
Usability is an issue that impacts us all the time. It cannot be ignored. Usability testing must be conducted frequently and carefully. Otherwise designers will have created a product that cannot be used by the very people it was designed for.
Lynne, Cooke (2005).Eye Tracking: How It Works and How It Relates to Usability. Technical Communication. 52, 456-463.
No comments:
Post a Comment