Having a Bad Day? Detecting the Impact of Atypical Events Using Wearable Sensors
Life events can dramatically affect our psychological state and work performance. Stress, for example, has been linked to professional dissatisfaction, increased anxiety, and workplace burnout. We therefore explore the impact of atypical positive and negative events on a number of psychological constructs through a longitudinal study of hospital and aerospace workers. We use causal analysis to demonstrate that positive life events increase positive affect, while negative events increase stress, anxiety and negative affect. While most events have a transient effect on psychological states, major negative events, like illness or attending a funeral, can reduce positive affect for multiple days. These findings provided motivation for us to train machine learning models that detect whether someone has a positive, negative, or generally atypical event. We show that wearable sensors paired with embedding-based learning models can be used “in the wild” to help detect atypical life events of workers across both datasets. Extensions of our results will offer opportunities to regulate the negative effects of life events through automated interventions based on physiological sensing.