PARIS, Aug 11 — New research published on Friday in the journal Science suggests we shouldn’t be so quick to accept positive internet reviews and ratings due to what its authors call “herding,” a phenomenon of consumers liking something simply because someone else likes it.

The study, co-authored by an MIT professor and conducted over a five-month period using a news aggregation site, found that when researchers randomly selected and artificially inflated the popularity ratings of certain user comments, their popularity continued to “snowball,” receiving a 25 per cent higher average rating from other site users.

Comments manipulated to have positive ratings were also 32 per cent more likely than untreated comments to receive a favorable rating from the very next viewer of those comments, and 30 per cent more likely than untreated comments to obtain a very high favorable rating. All of which suggests that we are heavily influenced by the positive opinions of others when expressed online.

Although the site used for conducting the study remains nameless for legal reasons, the research team, led by Lev Muchnik of the Hebrew University of Jerusalem, Sean Taylor of New York University, and Sinan Aral from MIT, did confirm that it was similar to Reddit, a popular site where users can leave comments on any article which other users can then up-vote or down-vote if they like or dislike it.

Advertisement

When comments were voted up, other users followed suit, something the researchers call herding—as in sheep.

“This herding behavior happens systematically on positive signals of quality and ratings,” says Aral, who also notes the results “were asymmetric between positive and negative herding.”

Comments given negative ratings attracted more negative judgments, but that increase was drowned out by what the researchers call a “correction effect” of additional positive responses. Aral claims that this is because people are more skeptical of negative social influence. “They’re more likely to ‘correct’ a negative vote and give it a positive vote,” he said.

Advertisement

Rating systems and the influence on web usersperceptions

The findings raise a serious issue regarding influence. These techniques could be used to favorably alter web users’ opinions on any number of subjects or products in order to change consumer behavior. “Real and important decisions are made based on these ratings and, in fact, these rating systems are a big part of consumers’ confidence in e-commerce transactions,” said Aral, “[Consumers] rely on them to judge the quality of products and services online.”

However, the study also found that there are limits to herding. Unlike articles related to politics; culture and society; and business, manipulated comment rankings under articles about economics; IT; and general news did not influence other users’ likelihood to up-vote.

In March, a study published by Dominique Brossard and Dietram A. Scheufele, two professors of science communication at the University of Wisconsin, Madison found that when examining the content and language of below the line comments on news websites and blogs, the worse the tone, the bigger the negative impact on readers.

They took 1183 volunteers and gave them an article from a fictional science blog to read surrounding a new technology.

The piece was balanced and weighed out the technology’s risks and benefits. The volunteers were randomly given one of two versions of the piece, the first featured natural or what the researchers called “civil” reader comments, while the second version contained aggressive and argumentative reader comments below the line.

The readers were then asked to complete a questionnaire about the article.

“The results were both surprising and disturbing,” Brossard and Scheufele wrote in a New York Times article about the experiment, “Uncivil comments not only polarized readers, but they often changed a participant’s interpretation of the news story itself.”

The authors dubbed this “the nasty effect” because they found that just a single comment detailing a personal attack on the writer rather than on the subject matter, was enough to make volunteers believe that the technology’s downside was greater than they had thought.

However, the “civil” group’s reaction to the article remained constant even though the ‘civil’ comments communicated the same agreements or disagreements, but in a much more acceptable and well-mannered tone.

The researchers’ greatest concern from the results was that the internet in general and blogs in particular are many people’s primary source of scientific information and therefore due to “the nasty effect” many consumers are unconsciously becoming misinformed about potentially serious issues and important concepts. – AFP-Relaxnews