According to a new study, LinkedIn conducted experiments on more than 20 million users over a five-year period that, while intended to improve how the platform worked for members, may have had an impact on some people livelihoods.
From 2015 to 2019, LinkedIn conducted experiments around the world that randomly varied the proportion of weak and strong contacts suggested by its "People You May Know" algorithm — the company automated system for recommending new connections to its users. Researchers from LinkedIn, the Massachusetts Institute of Technology, Stanford University, and Harvard Business School co-authored a study published this month in the journal Science.
LinkedIn algorithmic experiments may have caught millions of people off guard because the company did not notify users that the tests were taking place.
According to experts who study the societal effects of computing, conducting long, large-scale experiments on people that could affect their job prospects in ways they are unaware of raises concerns about industry transparency and research oversight.
"The findings suggest that some users had better access to job opportunities or that there was a meaningful difference in access to job opportunities," said Michael Zimmer, an associate professor of computer science and the director of Marquette University Center for Data, Ethics, and Society.
The Science study put to the test an influential sociology theory known as "the strength of weak ties," which holds that people are more likely to gain employment and other opportunities through arms-length acquaintances than through close friends.
The researchers investigated how changes in LinkedIn's algorithm affected users job mobility. They discovered that relatively weak social ties on LinkedIn were twice as effective as stronger social ties in securing employment.
LinkedIn stated in a statement that it "acted consistently with" the company user agreement, privacy policy, and member settings during the study. LinkedIn's privacy policy states that it uses personal data from members for research purposes. According to the statement, the company used "noninvasive" social science techniques to answer important research questions "without any experimentation on members."
The study goal was to "help people at scale," according to Karthik Rajkumar, an applied research scientist at LinkedIn and one of the study's co-authors. "No one was put at a disadvantage in their job search."