Online ‘Likes’ Herd Others to Similar Views, Study Finds

Photographer: Simon Dawson/Bloomberg

A Facebook Inc. "Like" logo sits on display at the company's new data storage center in Lulea, Sweden. Close

A Facebook Inc. "Like" logo sits on display at the company's new data storage center in Lulea, Sweden.

Close
Open
Photographer: Simon Dawson/Bloomberg

A Facebook Inc. "Like" logo sits on display at the company's new data storage center in Lulea, Sweden.

Positive opinions are more influential than negative ones, at least on the Internet.

If an article is “liked” on a website such as Facebook or Reddit, new readers are more likely to approve of it, according to a study published in the journal Science. While the positive reactions create a “herding” effect, the authors said, negative views don’t appear to affect people the same way.

Using an undisclosed news-aggregation website, the scientists tinkered with the favorability ratings of certain comments on the site. The comments that got a positive boost from the researchers subsequently took off in popularity, receiving a 25 percent higher rating on average from other users. In other words, people believe the hype -- at least some of the time. That’s not always a good thing, researchers said.

“If someone’s made a comment about a product I want to buy, I assume it’s useful if it’s been voted up and not useful if it’s voted down,” said Matthew O. Jackson, a professor of economics at Stanford University near Palo Alto, California, who wasn’t involved in the research. “So we care that the right things are pushed up. It’s a good sorting device for information, and if it’s not being done well, that’s bad news.”

Stories categorized as politics, culture and society and business generated the positive bandwagon response, while those under economics, IT, fun, and general news didn’t.

What’s more, when the site altered ratings in a negative direction, people were more skeptical, the study authors found, and were more likely to cancel out a negative vote with a positive vote.

Positive Reinforcement

“One possibility is that seeing something positive makes you feel better about seeing something positive, and if you see something negative, you react to try to bring it back to zero,” Jackson said in a telephone interview.

Researchers during the five-month study randomly altered the ratings of 101,000 comments. Those manipulated to be more positive were about one-third more likely than unaltered comments to receive a positive rating from the next viewer, and 30 percent more likely to achieve a high favorable rating.

The results published yesterday suggest a certain skepticism about using the collective judgment to evaluate the quality of products or ideas, the researchers said.

“These positive ratings also represent bias and inflation,” said Sinan Aral, a study author and associate professor at the Massachusetts Institute of Technology Sloan School of Management, in a statement. “The housing bubble was a spread of positivity, but when it burst, some people lost their savings and their houses went underwater.”

To contact the reporter on this story: Elizabeth Lopatto in San Francisco at elopatto@bloomberg.net

To contact the editor responsible for this story: Reg Gale at rgale5@bloomberg.net

Bloomberg reserves the right to remove comments but is under no obligation to do so, or to explain individual moderation decisions.

Please enable JavaScript to view the comments powered by Disqus.