Eye tracking is one of a number of user test tools with which to analyze numerous indicators on the functioning of the visual system of users.
There are many controversies around this technique. I would like to clarify a few points.
Let’s first of all have a look at how it works.
- The data provided by eye tracking do no more than simply indicate position x, y and so on and the diameter of the pupil. These data are gathered at a regular frequency. At the end of the test, one has a series of data x and y that correspond to a diameter of the pupil. These raw data provide information on the real movement of the eyes. They never allow you to infer behaviour.
- Based on these raw data, algorithms analyze the points x and y to extract so-called ocular fixations. An ocular fixation consists of a number of raw points that eye tracking has recorded. To put things simply, the duration of the fixation is obtained by the number of x and y points that are close to one another.
- After the treatment of the data, software analyzes the data from different points of view.
And that’s where the problem starts. People who lack a knowledge of the basics of the visual system and its use in our global behaviour, will only analyze the moving points on a screen or clouds of points that represent the average visualisation of users.
I’ve read the following on a usability post. And unfortunately, this represents the common practice:
“I have to admit I have discovered as much by analyzing the testers as by analyzing the eye tracking results (if not more, to be quite frank. But that may have something to do with my relative lack of experience?). Heat maps, however, that represent the zones of heat, which are the zones on the screen users have looked at for a longer period of time, are an undeniable plus. When talking with someone who only has a limited knowledge of the subject, they will help me to defend the point of view of usability in an irrefutable way.” (source : http://www.expressions.be/2007/10/07/tobii-eye-tracking-test/)
This kind of remark shows a lack of knowledge of the basics of the visual system. A pity because this means people will discard a precious tool without even realizing it, and worse, they will use material to impose their own vision of things, by merely stating their opinion is based on scientific fact.
When you are familiar with the fundamentals of human behaviour, eye tracking and a number of other techniques, it allows you to do a number of things. Some examples.
- Visualise hesitation patterns between different words in a navigation
- Detect whether a word or a text is read or not
- Detect the cognitive implication generated by each element of the interface
- Time-measure the discovery of different zones of the interface
- Measure the gaps in the eye movements as forecasted by the expert and those actually performed by the users
- Know in which lobe of the brain the information, captured by the eye, is analyzed.
All these data, consolidated with other behavioural indicators, allow you to modify interfaces precisely and objectively.