Women use lots of exclamation points in tweets about yogurt and yoga, while men like to talk about Jeeps and Vegas.
Researchers at the Mitre Corporation used a machine to figure out a person’s gender just by reading their tweets. The program correctly guessed peoples’ gender 76 percent of the time. And, after evaluating only one tweet, the computer was able to predict the correct gender 66 percent of the time.
Mitre’s experiment featured two study groups — the first one featured tweets from people whose gender was known. The second, unknown. In addition to tweets, researchers collected some profile data like screen name, full name, location, URL and description.
Mitre found that women tend to use words like “mom,” “yummy,” “wait,” “wanna,” “chocolate” and “hair,” in addition to extended vowels, such as “soooooo.” Men use terms like “google” and “http.”
Researchers, which also identified words that followed “my,” concluded women tend to use “my bff,” “my yogurt” and “my research.” Guys say “my zipper,” “my jeep” and “my gf.”
The study also found a political bias. Democrats are more likely to use words like “upscale,” “yoga” and “sushi,” while Republicans mention “weapons,” “blackberry” and “marine.”
Of course there is no hard-and-fast rule here, but the researchers found if a tweet does contain gender or political-specific words, it’s a good bet they represent a gender and political perspective.
The experiment tested the different speech patterns of men and women — certain words, phrases and even punctuation — to see if it extended into cyberspace. Sociolinguists have known women laugh more than men, but they wanted to see if it translates into more exclamation points in tweets.
The idea that we are what we tweet isn’t new. Research has been looking at Twitter to look for general trends. Last May, a report indicated Wall Street is looking at Twitter to help make bette investment decisions.
At Indiana University, algorithms rated millions of tweets to determine an overall mood of the country, hoping to predict the direction of the Dow, since general moods are widely believed to play a role stock prices.
The implications for marketing, politicians, and even financial markets will continue to evolve and transform as more people pick up on social media to communicate — and to do so often on more levels than intended.