Key To Viral Posts: Naming Political Opponents

Social media users are more likely to share posts that criticize political opponents by name than those which praise people and ideas they support according to newly-published research. The pattern was consistent across different sites and people with different views.

Researchers at the University of Cambridge were exploring what makes some posts on Facebook and Twitter more likely to be shared. Previous research had suggested the main driver was the tone of the language, with negative emotions such as anger, or a "sense of moral indignation" the most likely to lead to a post being widely shared.

The Cambridge researchers looked at posts from US news sites (including those viewed as having liberal and conservative slants) and posts from US congress members of both main partners. They then analyzed the language in the posts and how many times they had been shared. (Source: cam.ac.uk)

Ingroups And Outgroups Key

As expected, tone certainly had an effect. They found every negative word increased the chances of a post being shared by 14 percent, while every positive word reduced the chances of a share by five percent.

However, the researchers also examined whether the post was referring to a political "ingroup" or "outgroup". The ingroup was where the subject of the post was from the same political side as the politician making it, or associated with the political slant of the newspaper. The outgroup was the opposite, for example a Republican Senator posting something critical about a Democrat or a liberal newspaper posting something negative about a Republican.

The researchers found a post being about the ingroup made no significant difference to a post being shared. However, every word in a post that referred to or identified an outgroup member increased the chances of a share by 67 percent.

Algorithms Start Vicious Cycle

In other words, using terms such as "Biden", "Trump", "Democrats" or "Republicans" multiple times made more difference than how many negative words the post used about them. Perhaps the cynical view is that users see the names of opposing parties and politicians as the most negative words of all.

The researchers said this pattern risks becoming a self-perpetuating cycle because of algorithms that means users are more likely to see posts that have already been widely shared. They also suggested the findings mean avoiding the "echo chamber" problem might not be as simple as it seems. (Source: bbc.co.uk)

That's because even if somebody starts looking at content from "the other side" of the political divide, the shares-based algorithms mean they are more likely to see criticism of their own side than a positive case for the beliefs, arguments and policies of the other side.

What's Your Opinion?

Are you surprised by the findings? Is it possible - or desirable - to expose people to a wider range of argument on social media? Is it dangerous to simply use the number of shares as a way to decide which posts people should see first?