One’s circle of friends might assist them to get a better interpretation about their overall wellness and health than merely utilizing wearable devices like Fitbit, as per scientists, comprising one of Indian origin. The research, issued in the PLOS ONE journal, examined what the makeup of social networks states about the state of happiness, health, and stress.
A Professor from the University of Notre Dame in the United States, Nitesh V Chawla, said, “We were fascinated about the social network’s topology—what does my place in my social network forecast about my well-being and health? What we discovered was the social network structure offers a noteworthy enhancement in the predictability of health states of a person compared to the just utilizing the information obtained from wearables, such as the heart rate or the number of steps.”
For the research, the study participants wore Fitbits to record the health behavior data—like sleep, steps, activity level, and heart rate—and accomplished self-assessments and surveys about their feelings of happiness, positivity, and stress. Chawla and his associates then examined and modeled the information, utilizing machine learning, together with social network features of an individual, comprising centrality, degree, number of triangles, and clustering coefficient.
These features are indicative of properties such as social balance, connectivity, closeness, and reciprocity within the social network. The research displayed a strong link between social network structures, the number of steps, level of activity, and heart rate. Social network structure offered noteworthy enhancement in forecasting one’s well-being and health in comparison to just examining health behavior records from the Fitbit alone.
Likewise, the language of people could disclose hints about their future threat of developing psychosis. Scientists from Harvard University in Boston, MA, and Emory University in Atlanta, GA, concluded this after examining the subtle characteristics of people’s daily speech utilizing a machine-learning method.