To understand this, the team employed a new methodology that used Google News service which detects which "trending topics" appear in traditional media sources.
"Approximately half of the events included in 'trending topics' are also reported as news in the traditional media, while the other half are subjects that, despite attracting the attention of a large number of Twitter users, do not appear in the media," the team noted.
The study also analysed who reports the news first, comparing the publication date of the "trending topics" on Twitter with the related stories that run in the country's main dailies.
"If we look at the news that is reported by both sources, more than 60 percent of it appears first on Twitter, while less than 10 percent appears first in the traditional media (the rest usually appears the same day)," the researchers remarked.
That does not mean, they explained, that there is a "tweet" that contains news, but that the subject attracts the attention of a certain number of users to qualify for the category of "trending topic" on the social network.
The analysis focused on the "trending topics" of Twitter because they share some of the same characteristics as news, dealing with subjects that attract the attention of a large number of people.
"They are events that a large number of users are interested in and, in this regard, we can say that they are news items selected democratically by Twitter users in a country," the researchers noted.
For the study, all the "trending topics" on Twitter were compiled from 35 countries over three months in 2013 and from over 62 countries over the same period in 2014.
In total, more than 300,000 "trending topics" generated in different countries and at different times were obtained.
"We found that the geographic dissemination of news on social networks preserves some of the biases present in the dissemination of traditional news, like the fact that it tends to flow more from rich countries to poor countries," explained Ruben Cuevas, researcher from Universidad Carlos III de Madrid (UC3M).
The study was published in the journal PLOS ONE.