Researchers from ETH Zurich have developed a model that identifies
To develop your own stress model,The researchers observed 90 participants in the experiment in the laboratory. They were asked to perform normal office tasks that were as close to real-life conditions as possible. For example, tasks included analyzing various data, planning and coordinating meetings.
The researchers recorded data on howparticipants manipulated the mouse and typed on the keyboard. In addition, they collected data on heart rate (HR) and periodically conducted a survey to determine the level of stress.
The control group was given conditions to work withouthindrances, to others - on the contrary, they constantly created various problems. For example, work was constantly interrupted by chat messages, or participants were asked to urgently break away and take part in an interview.
The results of the experiment showed that peoplewho are under stress, control the mouse and work with the keyboard in a different way. The accuracy of moving the cursor decreases, and it travels a greater distance on the screen. In contrast, people in a calm state move the mouse smoothly along the shortest distance to the desired location. What's more, people who are stressed in the office make more typing mistakes. They write in fits and starts, making many short pauses.
The researchers note that the results of the model,formed as a result of training on the data collected during the experiment predict stress in workers better than heart rate analysis. They believe that the data from this model can be used for workplace monitoring and timely detection of dangerous symptoms. But first, we need to work out a data protection system that will ensure the anonymity of employees.
People will accept and use our technologyonly if we can guarantee the anonymity and protection of their data. We want to help employees identify stress at an early stage, not create a monitoring tool for companies.
Jasmine Kerr, psychologist and study co-author
Read more:
Look at the highest resolution map of Mars: 110,000 frames and 5.7 trillion pixels
"Sea" of quarks inside one proton: what does an elementary particle consist of
New image of Hubble puzzled scientists