The eyes have it
Matteo Penzo has written a very nice introduction to eyetracking in usability studies over at UX Matters.
The only thing I’m not entirely convinced about in the article is the idea that you can adopt a think aloud protocol and measure eye movements at the same time. Think aloud already has its own set of limitations (e.g. it’s a relatively unnatural scenario for the user, it’s full of interruptions and there’s the potential for subconscious bias and “priming” by the tester) so there’s a risk that you might end up measuring eyetracking artefacts that don’t occur under more natural user scenarios. The two methods are designed to evaluate very different aspects of user behaviour, so I’d always be inclined to keep them separate and play to their strengths. i.e using think aloud to evaluate and understand user expectations and mental models, and eyetracking to get quantitative data in specific task-based situations. Of course, there’s no reason why you couldn’t use both approaches in separate tasks within a test session to get a broad picture of user behaviour (assuming that’s what you were looking for!).
There are more articles on eyetracking in the pipeline for future issues of UXMatters, which should be well worth checking out.
Comments
2 Responses to “The eyes have it”
Leave a Reply
About
I'm Stuart Church, a user experience consultant with Pure Usability in Bristol, UK. Sensorydrive is my personal blog and covers user experience design, information architecture, product design, psychology, research methods, perception and pretty much anything else that takes my fancy! You can find out a bit more about me if you want...
Latest Flickr images
Hi Stuart and thanks for pointing to my E.T. article.
I think or – better – I’ve experimented that eyetracking and think-aloud can go together. You might want to use a simpler think-aloud protocol (since it’s obvious to the eyetracker what the user is looking at even though she doesn’t mention it in her think-aloud).
We use to both (i) videotape and (ii) gaze record the user behaviours (i) and interactions (ii) in a way that makes possible for us to have a deep quantitative (mainly ii) and qualitative (mainly i) analysis.
So – for example – we can say that while the user was upset (i) for not finding the “search” field she actually looked at it (ii) several times with no understanding of the field functions. This kind of data merge would be impossible without the syncronization of eyetracking test and think-aloud.
Just my humble opinion.
Hi Matteo
Thanks for the feedback and also for being the first person to comment on my new blog!
I understand what you’re saying about the kind of data merge you can get with the synchronization of think-aloud and eyetracking, and I do agree that it looks like a useful approach to take. My point really is that are often likely to be trade-offs when you combine two approaches. For example, you’d never attempt quantify task completion time in a think-aloud protocol because the constant human intervention interrupts the task. Thus, you get think-aloud data at the expense of quantitative data. The same must be true to a degree with eyetracking studies – if you interact with the test subject, you’re not really quantifying the same thing as you would were they left to do the task on their own. This doesn’t mean that one approach is necessarily better than the other; it just depends on what the question is that you want to answer with your user testing. In some cases, a mixture of eyetracking and think-aloud may be best approach to take. In others, pure eyetracking studies in a more formal test scenario may be the right way to get the data you need.