Press: Marcel Schliebs in the FT on TikTok’s battles to stay ‘apolitical’ ahead of the US election
27 October 2020
The team’s research on junk news during the 2019 Indian Elections has been widely covered, including in the following:
Hindustan Times: SP, BSP’s election posts least sensational, says Oxford study
The analysis of data between February 14 and April 10 suggests more than a quarter of the BJP’s content and a fifth of the Congress’s content is ‘junk news’, while the SP and BSP share ‘very little sensational, extremist, or conspiratorial content’.
The analysis as part of the Oxford Internet Institute’s Computational Propaganda Project looked at WhatsApp, Twitter and Facebook posts during the period that included the Pulwama attack, India’s airstrike in Pakistan and related events…
Narayanan said: “Our study, using significant volumes of data from Facebook and WhatsApp, shows junk news and misinformation has been widespread in the Indian election campaign.Even if this content has not been generated by party workers themselves, the material has been widely shared by party supporters on social media”.
“While it is difficult to say how much this will influence the outcome of the 2019 Indian elections, it is clear that with many Indian people moving away from traditional print media to online sources, there are huge questions around the influence of technology and social media platforms in modern day democracy.”
A study by the University of Oxford revealed this week that ahead of the vote, more than a quarter of news links shared on Facebook pages supportive of Modi’s ruling BJP party “were classed as junk news and information.”
More than a third of images shared in BJP WhatsApp groups were “classed as divisive and conspiratorial.” For the Congress Party, the figures were 21 per cent and 28.5 per cent respectively. Requests for interviews to both parties remained unanswered.
Some of this material is passed on not only from partisans, but from the main parties as well.
With a dearth of academic research regarding WhatsApp, the researchers analysed 27,000 posts from 130 public Facebook pages and over 200 public WhatsApp groups between February 14 and April 10.
Underneath a broader umbrella of “divisive and conspiratorial content,” the researchers have coined the phrase “junk news.” They determined the “junk news” label based on professionalism, style, credibility, bias and counterfeit content. The researchers consciously avoided the phrase “fake news” because of how leaders such as Donald Trump have “co-opted” the phrase, Narayanan said.
27 October 2020
23 March 2020
21 May 2019
13 May 2019