Researchers from the Swiss Federal Institute for Technological Research (ETHZ) have succeeded in designing a new machine learning algorithm that can estimate the stress level based on the way the mouse clicks and the keyboard is typed. With the help of this achievement, it is possible to detect the high level of stress in people and to prevent the occurrence of chronic stress in work environments.
What does typing and using a mouse have to do with stress?
Researchers at the Swiss Federal Institute for Technological Research (ETHZ) have announced that they have succeeded in designing a new model with the help of machine learning that can estimate people’s stress levels based on how they interact with the mouse and keyboard. They claim that the way we use electronic devices, especially in work environments, is closely related to the level of stress and heart rate.
To design this model, 90 participants in a study were observed to perform tasks similar to administrative tasks such as setting work appointments or analyzing data. During the research, the way of using the mouse and keyboard was recorded along with the heart rate of the subjects, and questions about the volunteers’ feelings were regularly asked.
While some participants were allowed to continue their work without disturbance, others were faced with disturbing factors such as messages or attending job interviews. Finally, it was found that people under stress move their mouse in a different way or their typing style is different from the first group. This change was such that the stressed volunteers moved their mouse more often and less accurately, and their mouse pointer covered a wider area of the screen. Also, the number of typing mistakes of these people is more and they are used to having more but shorter stops while typing a sentence.
The phenomenon observed in the recent study can be justified by the theory of neuromotor noise; The hypothesis according to which it is said that as the level of stress increases, the brain’s ability to process information decreases. This complication can lead to reduced movement abilities. Researchers believe that using the new findings, it is possible to identify employees who are under high stress in the workplace faster and easier than before.
Researchers say that in many cases, people under stress are not aware of the decrease in physical and mental ability in addition to their productivity, or they find out about this very late. It is said that the new model is being tested on Swiss employees who have agreed to have their heart rate and mouse and keyboard usage recorded through an app. More results are expected to be published by the end of this year.
The researchers point out that they are aware of the sensitivity of the collected data, and for this reason, they will be stored and reviewed anonymously and with high accuracy. Also, their goal is not to create a monitoring tool for companies and governments, but they only want to improve the health level of employees.