Ladies and gentlemen, you may officially toss out your collection of mood rings. Now there’s a better way to check if we’re happy or not: Twitter.
Computer scientists at Northeastern University have just released a study which presents conclusions regarding the relative happiness of America based on “sentiment analysis” performed on tweet content. One key finding that’s making headlines and ruffling feathers is that Americans living on the West Coast are happier than those living on the East Coast.
For the study, the scientists at Northeastern performed the content analysis on nearly 300 million tweets from Americans and then indexed them according to time of day, sentiment, and location. The analysis was performed using the Affective Norms for English Words, or ANEW, system developed at the NIMH Center for Emotion and Attention at the University of Florida. As described in its manual, ANEW was developed “in order to provide standardized materials that are available to researchers in the study of emotion and attention.” Basically the ANEW system allows researchers to assign different values to a set of pre-selected words in order to determine a relative scale of emotions.
However, as is often the case with an attempt to project standardized values on subjective human behavior, ANEW is not completely reliable. As one article discussing the paper’s findings noted, “if someone types tweets ‘I am not happy’, the system counts the tweet as positive because of the word ‘happy’.” That being said, even as an imperfect tool, the practice of “sentiment analysis” using the ANEW guidelines is actually gaining popularity among large corporations as a tool to measure brand awareness and reactions.
Unfortunately for the study conducted by Northeastern University, most of the results aren’t exactly earth shattering revelations. For example, apparently most of us hate our jobs and prefer the weekends. But the video they put together does offer a nice visual of how the happiness or relative gloominess progresses over the course of an average workday.
What do you think? Do you think text-based “sentiment analysis” is a legitimate form of making these types of conclusions? Do you think the West Coast is, in fact, happier than the East?