1013378_laptop_magnifying_glass_search_iStock_000010540776XSmall

A quick “social media” news search reveals that social media use is on the minds of a number of political parties. From Berkely City Council, to the US Senate, the ways in which social media can be used to engage community members, influence opinions, and predict voting polls has become a top priority. However, new research from Fei Xiong and Yun Liu reveal that the influential power of social media might not be all that it promises to be.

Xiong and Liu (2014) investigated how opinions form and change on Twitter. Their study, which looked at 6 million Tweets from December 2010 to June 2011, had two main findings:

  1. “Public opinion often evolves to an ordered state in which one opinion predominates, but not to complete consensus”
  2. “Agents are reluctant to change their opinions, and the distribution of the number of individual opinion changes follows a power law.”

In addition, Xiong and Liu highlighted that even if opinions did change, users would not necessarily admit to this, nor can we be sure whether or not they are telling the truth at any point.

Although some people use their “real” names as their username, the Twittersphere is somewhat anonymous in that you don’t know the personality traits, mood, or previous opinions of the person you are conversing with, unless you know them outside their active Twitter persona. As such, users opinions are latent until someone Tweets/takes action. However, opinions may change, but no action (a secondary Tweet) is made to confirm this change. In turn, this dynamic affects the users followers and their actions and/or latent opinions.

These changes of opinion, latent opinions, or false opinions can be very confusing to organisations and Government’s concerned with public opinion and are known to lead them astray. Within Australian politics alone, opinion polls have failed to predict four of the past seven elections.

The study is groundbreaking in that it is one of the first to apply empirical evidence to opinion dynamics in the social media sector. Also, it raises the questions:

Is it possible for organisations to ever really identify users’ true opinions and opinion changes online?

Is there potential for a new method of influencing opinions of social media/Twitter and if so, what would it look like?