Hacking around on the keyboard and sending the curser on long journeys using a computer mouse – this can be a sign of stress in everyday working life. According to a Swiss study, typing on the keyboard and moving a mouse can better indicate how stressed a person feels than the heart rate otherwise measured for stress proofs. Mathematicians Mara Nägelin and colleagues from the Swiss Federal Institute of Technology (ETH) Zurich present their results in the journal "Cell".
"If you are stressed, you move the mouse pointer more often and less accurately and travel longer distances on the screen. Relaxed people, on the other hand, reach their destination on shorter, more direct routes and take more time," says Nägelin. Stressed people made more typing mistakes and they wrote more choppy, with many short pauses. Relaxed people took fewer, but longer breaks when writing. How can the link between stress and typing and mouse behavior be explained? "Increased stress negatively affects our brain's ability to process information. This also affects our motor skills," said psychologist and co-author Jasmine Kerr (also ETH).
Typing and mouse behavior show stress better than heart rates
The psychiatrist and stress researcher Mazda Adli, chief physician of the Fliedner Klinik Berlin, described the methodology of the study as groundbreaking. "This is an interesting approach to the investigation of individual stress susceptibility," he told the German Press Agency. "In the future, the method could be used to investigate for oneself how susceptible to stress and faults one is under certain external conditions, then change something in the environment and see whether the susceptibility to stress has changed." He was not involved in the study.
Mouse and keyboard behavior as well as heart rates of 90 people were recorded. All of them completed realistic office tasks in the laboratory. Some remained undisturbed, others also went through a job interview or constantly received new chat messages. The scientists used machine learning and asked people about their perception of stress. "We were surprised that typing and mouse behavior predicts how stressed subjects feel better than heart rate," says Nägelin.
In everyday working life, many people felt distracted by constantly new emails, chat messages or phone calls, Adli said. "Distractibility triggers stress." Then it makes sense to shield yourself against disturbing stimuli: for example, read e-mails only every two hours, signal if you do not want to be disturbed, or take regular breaks. Adli points out that not all stress is negative. A public lecture or competition can create a peak of stress that can be stimulating and provide good performance or even a pleasant feeling. It becomes problematic when stress peaks do not subside and those affected can no longer recover from them.
Whether stress detection according to the ETH model makes sense in the workplace, for example to prevent damage to health, is a delicate question. "We want to help workers detect stress early, not create a monitoring tool for companies," Kerr said. Adli can imagine the use more for his own assessment. From an occupational medical point of view, an assignment would only be conceivable under absolute anonymity, according to Adli.