Researchers at Brigham Young University may have found a next-gen way of measuring user behavior on the web.
Work began as a means to test the theory that users use a mouse differently when their mood is altered.
Lead researcher, Jeffrey Jenkins and his team set out to prove the theory by evoking anger from their subjects and then tracking their mouse movements.
To rile the subjects, Jenkins asked each to take a test that was purported to measure intelligence, but was purposefully set up to perform slowly and penalise users for giving incorrect answers. He then informed the subjects that low scores equaled low intelligence.
The research identified that frustrated, sad or anxious people are more likely to use the mouse in a jerky and sudden manner, but in a surprisingly slow fashion. The mouse movements, as a result, tend to be less precise. People who feel frustrated, confused or sad are less precise in their mouse movements and move it at different speeds.
Jenkins said: “It’s counter-intuitive; people might think, ‘When I’m frustrated I start using the mouse faster.’
“Well, no, you actually start moving slower.”
Jenkins expects his research will carry over into applications for developers – allowing them to learn how users are interacting with sites. This knowledge could be anything from knowing when users are frustrated with content, to understanding the time limit for becoming so frustrated that a user abandons the process.
He also expects that the research will yield similar results on mobile, but the theory has yet to be tested thoroughly.
This kind of data could lead to a better understanding of user frustrations in order to improve site UI/UX. It could also lead to behavior-based targeting as well as specialised actions for specific behaviour groups.
The findings uncovered could open the door for the next level of advanced web analytics. As big data, machine learning and advanced AI continue to make big strides, it’s only a matter of time before we find other, more assured ways of gathering critical data on user behavior.