LinkedIn Defends Testing On Users

LinkedIn has admitted it ran experiments that may have affected a user's chances of getting a job. The business networking site says everything it did was covered by its user agreement.

The New York Times revealed that the experiments involved more than 20 million users over five years. The core of the testing was the way LinkedIn picked who appeared in the "People You May Know" section that automatically suggests potential contacts.

The experiment, run alongside three major universities and later published in the journal Science, meant some users got a higher proportion of suggested contacts which were likely to have strong social ties to the user. Others got a higher proportion of suggested contacts with likely weaker social ties. The study also intentionally varied the number of suggestions each user saw.

The researchers then studied how successful the different users were at finding employment during the test period. It turned out that those who were suggested contacts with weaker social ties were twice as likely to get jobs as those who'd been suggested contacts with stronger social ties.

Weak Ties Expand Opportunities

Though the results may sound counterintuitive, they actually supported the hypothesis that the researchers were trying to test. The researchers believe weak social ties help people connect with those wider range of cultures, locations and business types, increasing the chances of matching a suitable candidate with a job.

Contrastingly, strong social ties tend to be with people from similar backgrounds, making it less likely somebody will discover an opportunity they didn't already know about. (Source: science.org)

Ethics In Question

The study has proven controversial, however. Academics and ethics experts say the study itself is sound, but question the lack of informed consent among the participants. That leads to fears that LinkedIn's actions may have affected a user's career opportunities.

LinkedIn insists it had valid consent through its user agreement, which gives a broad warning that it could use customer data for research. Critics say this wasn't sufficient for research that involved actively changing the user's experience on the site and could have real life consequences. They say LinkedIn should have sought explicit consent to carry out this research, explaining what it involved and how it affected users. (Source: arstechnica.com)

What's Your Opinion?

Was LinkedIn right to carry out this research? Do you assume suggested contacts on social media and networking sites are always decided in the same way? Would you have consented to your account being part of such research?