Exercise Contagion on an Online Social Network
https://www.nature.com/articles/ncomms14753
The article above details a study conducted by Aral and Nicolaides on the contagion effects of exercise on an online social network. By examining data on network ties and daily exercise patterns of 1.1M individuals who ran over 350M kilometers over 5 years, the authors add to our understanding of cascading behavior across social networks. The study’s main finding was that exercise is socially contagious, when mediated by online social networks that facilitate sharing of exercise data (such as the distance that a person runs on a given day). Given the popularity of exercise-related social networks like Strava and Fitbit, this study offers us relevant insights on how these networks can shape our everyday behaviors.
The results showed strong exercise contagion efforts—on the same day, on average, an additional kilometer per minute run by friends influenced an ego to run at additional 3/10th of a kilometer per minute faster, while an additional 10 calories burned by friends influenced an ego to burn 3.5 additional calories. While this is a “positive” behavior cascade, a “negative” behavior cascade also exists. For instance, using local weather data, the authors found that a rainy day in Nashville dampened a runner’s enthusiasm to run in Nashville and reduced his running performance, which in turn reduced the motivation for a runner in Chicago to run regardless of Chicago’s weather. It can be argued that there are direct benefits from imitating a friend’s running behavior—the benefits of running would be limited to personal health if none of your peers run; but if many of your peers run, you derive satisfaction and social status from running to compete with them.
The study found that the network structure around a node matters greatly in facilitating exercise contagion. In particular, the embeddedness of a relationship—the number of mutual friends between two people—can promote contagion due to the social monitoring in the network. For instance, when two people have many mutual friends (high embeddedness), there are more social sanctions and reputational consequences for not keeping up with one’s promises to stay fit. Simultaneously, there are greater social rewards for keeping up with one’s running regime or beating an exercise record, as one’s accomplishment would be valued highly by more friends.