Researchers at ETH Zürich, Switzerland, have developed a model to detect the levels of stress someone is experiencing at work based on how they type and move their mouse. The results may help prevent chronic stress before it becomes too serious.
In Switzerland, a third of workers experience workplace stress. In the US the figure is higher, with 40 percent of employees claiming their jobs are very or extremely stressful, while 25 percent believe their work is the most significant stressor in their lives. The arrival of the recent global COVID-19 pandemic made this situation worse, as its impacts on work and social environments has led to a marked increase in mental stress and related depression symptoms.
Those suffering from such stress often do not realize how bad the situation is until it’s too late and their physical and mental resources are spent. As such, it is important to identify work-related stress as early as possible so that it can be dealt with effectively.
The new model developed by Mara Nägelin, a mathematician and lead author of the study, and colleagues represents a valuable step in this effort. Using machine learning and new data, the model can detect stress levels at work based on the way people use their mouse and keyboard.
"How we type on our keyboard and move our mouse seems to be a better predictor of how stressed we feel in an office environment than our heart rate," Nägelin said in a statement.
To develop their model, the team observed 90 study participants in a lab as they performed simulated office tasks that were as realistic as possible, such as planning appointments or recording and assessing data. The researchers recorded the participant’s mouse and keyboard behaviors, as well as their heart rates. They also frequently asked participants to show stressed they were feeling.
The study participants were separated into two groups. Half were allowed to work undisturbed, while the other half were subjected to repeated chat messages and were asked to take part in a job interview. Unlike other studies into work stress where those in the control group were allowed to just relax and not undertake any work-based tasks, all the participants in this study performed office activities.
The results showed that the more stressed an individual is, the more erratic and imprecise their mouse and keyboard behavior was. “People who are stressed move the mouse pointer more often and less precisely and cover longer distances on the screen,” Nägelin said. “Relaxed people, on the other hand, take shorter, more direct routes to reach their destination and take more time doing so.”
Stressed people are more likely to make mistakes when typing, and they write in fits and starts with frequent brief pauses. In contrast, relaxed people take fewer pauses, but they tend to last longer when typing on a keyboard.
The connection between stress and these activities can be explained by something called the neuromotor noise theory. “Increased levels of stress negatively impact our brain’s ability to process information,” added Jasmine Kerr, a psychologist and coauthor of the study. “This also affects our motor skills.”
Interestingly, the team found that there did not seem to be too much variation in the heart rates of those in either the control group or the participants who were working with constant interruptions.
“We were surprised that typing and mouse behaviour was a better predictor of how stressed subjects felt better than heart rate,” Nägelin stated. This could be because everyone was taking part in workplace tasks, which seemed to mean their heart rates did not vary all that much.
The researchers are currently testing their model with data from Swiss employees who have agreed to have their mouse and keyboard activities monitored along with their heart rates. This will be done on the job using an app that also asks the participant to assess their levels of stress. The hope is that the new results will be available by the end of the year.
This work raises some difficult challenges, however. “The only way people will accept and use our technology is if we can guarantee that we will anonymise and protect their data. We want to help workers to identify stress early, not create a monitoring tool for companies,” Kerr added.
More research is currently exploring which features an app needs to satisfy this distinction so that sensitive data can be gathered and handled in a reasonable and ethical way.
The study is published in the Journal of Biomedical Informatics.