Major upheavals in people’s lives – whether positive (e.g., marriage, childbirth) or negative (e.g., cancer, natural disasters) - are often transformative. Sometimes the same events cause people to become more generous and virtuous, while other times they may exhibit cruelty and a host of vices. While the bulk of studies on transformative events has relied on self-reports, very little research has leveraged big data to understand how such events affect people’s lives. The goal of this project is to develop novel big data models that will allow us to track changes in positive and negative value-related behaviors in relation to major events. We specifically aim to: (1) build a large, diverse dataset of linguistic accounts of upheavals in people’s lives, covering typical events in one’s life, health-related events, and major natural events; (2) develop computational linguistic models to track temporal shifts in value-related behaviors, targeting a set of positive (virtues) and negative (vices) behaviors; and (3) understand the social dynamics of behavior changes by measuring demographic differences in order to better predict and ultimately shape how people respond to major events in healthy ways.
The project is expected to have far reaching implications. Our big data approach will allow us to track the ways people are expressing themselves on a daily basis. We anticipate learning how people from different backgrounds and cultures organize their worlds and respond to different types of challenges. By identifying aspects of their social lives, we can begin to make predictions about ways to improve the course of people’s lives after upheavals to optimize virtues and minimize adverse behaviors. Moreover, the resources that will be generated in this project (datasets and software tools) will be made publicly available, thus enabling future research as well as educational projects concerned with tracking and understanding the effects of major events on people’s lives.