Want to Get Published in the Times? Use Twitter in a Novel Way.
A recent study released by our very own Michael Macy, a professor of sociology and information science, utilized more than half a billion tweets from more than two million Twitter users to determine a correlation between mood and times of the week. To accomplish this task, tweets were analyzed by tagging certain words, such as “happy”, “excited” and “awesome” as positive, while words such as “angry”, “frustrated” and “afraid” received a negative tag. These +/- data points were mapped along an axis of time, on scales ranging from hourly to ones that would cover a full year.
While the information is mostly intuitive- we are happier on weekends, and dreary winter days are depressing; the underlying method is what is truly groundbreaking. A databank of this size is very relevant to the sociological community, as, historically, studies of this nature have been carried out using a much smaller sample size, often relying on questionnaires or surveys. By using a self-generating sample medium such as twitter new doors are opened to researchers to further understand human beings.
There are still some kinks to be worked out as this is, after-all, a study of the moods of twitter users- a subculture of people which is not necessarily proportional to the greater populace. But, as a checking mechanism to match previous studies and data, it has already itself useful- the trends found by this study match those of previous studies.
Pretty, attention-grabbing picture:
Sources: