A Tufts professor’s research shows how to work the virtual room
You’ve got all those online “friends”—some close, some you vaguely know, and some, frankly, you’ve never met. Which of these folks will help you get a job when you need it?
That’s the question Tufts assistant professor of economics Laura K. Gee wanted to answer using big data. Gee was completing her Ph.D. in economics at the University of California at San Diego when a chat with political scientist James Fowler inspired her to use Facebook data to test the results of a famous 1973 study by Stanford’s Mark Granovetter called “The Strength of Weak Ties.”
Granovetter’s 1973 theory, based on interviews with 100 people in the Boston area, concluded that in professional, technical and managerial jobs, weak ties tap into a broader network, making them especially valuable when you’re seeking new opportunities. Did Granovetter’s thesis hold up using large data? Gee discovered that yes, it did, but with a twist. Paradoxically, according to her research, a single strong tie is more likely to help you than a single weak tie.
Gee identified three Facebook indicators of relationship strength: tagging someone, which suggests that you frequently find yourself with that person in the same place at the same time; having many friends in common with someone, which suggests that you share more than just an interest in cat videos; and regular posting on someone’s wall, which indicates that you’re thinking about that person regularly. She used these tools to assign a friendship-strength number—from one to 100—to every “friend.”
Next, she had to determine whether a friendship led to a job. She and her collaborators, Jason Jones at SUNY Stonybrook and Moira Burke at Facebook, decided that if a person started working at a friend’s workplace within a year, then chances were that the friend had advocated on the subject’s behalf.
Finally, she took a random sampling of 1,200 people who had used Facebook in 2012 and had started a job at a friend’s workplace. That gave her nearly a million connections to work with. She analyzed the data to determine whether the connection that had led to each user’s job was strong or weak compared with others in that user’s network.
Analyzing her findings, Gee says, “More than 90 percent of ties are really weak on Facebook by our measures, so part of the reason people get jobs through weaker ties is because you just have a lot of weaker ties. Collectively, weaker ties are really helpful.”
Then came the surprise. When Gee tried to predict which type of tie Facebook users got jobs from, she got seemingly contradictory information. “We said, let’s try to assign a probability of the likelihood that the subject is going to end up working with a friend as a function of how closely tied they are. We just plopped the probability on top of all your friends’ heads that you’ll end up working at the same firm as them.”
The results showed that the chances of eventually working with a friend increase with tie strength. “That says that even though collectively weak ties are really important, individually a single stronger tie is more important than a single weaker tie. Sort of paradoxical, but it totally makes sense if you think about your own networking: You’re personally more likely to be helped by a stronger tie.”
Gee’s research suggests that the best way to land a new position is to broadcast your job search to your entire network, thereby mobilizing your small army of good friends, relatives and that guy you met at a party three years ago. But if you can only send the message to one person, she says, tap into someone near and dear to you. Though her research doesn’t answer the why, she theorizes that the better friends know you, the better equipped they are to sing your praises in a unique way, and the more motivated they’ll be to help you.
The study, which the Journal of Labor Economics ran online earlier this year, got plenty of media attention, which Gee found rewarding. “As an academic, I feel like oftentimes you do this research, it’s a lot of blood, sweat and tears, and then it gets published and eight people read it. It’s nice when you’ve got a paper that appeals to a little wider audience.”
Rachel Slade is a Boston-based freelance writer.