![Sample our Social Sciences journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/110822/TOC-Subject-Sample-2021-Social-Sciences.jpg"})
Abstract
In 2006, the Guardian opened many of its articles to readers’ comments to encourage a “conversation” between journalists and their readers. Readers responded enthusiastically, and by 2016 they had posted 70 million comments on the site. However, from the outset many journalists complained about the quality and tone of comments. Female and BAME (Black, Asian and minority ethnic) journalists in particular complained that they were subject to more abuse than their male, white counterparts. This study finds prima facie evidence to support the journalists’ claims. Using comments that had been blocked by moderators as a proxy for abuse and dismissive trolling, it was found that articles written by women did attract a higher percentage of blocked comments than those written by men, regardless of the subject of the article; this effect was heightened when the articles ran in a particularly male-dominated section of the site. There was also evidence that articles written by BAME writers attracted disproportionate levels of blocked comments, even though the research was not designed to reveal this. Preliminary research findings were published in the Guardian and readers were invited to comment on them. Guardian journalists’ experiences of comments were also surveyed. Both sets of responses are analysed here, in order to explore the contested nature of online abuse in an online news media environment, and to evaluate the potential of comments to “democratise” journalism.
Acknowledgments
The analysis of the comment data would have been impossible without the collaboration of the Guardian’s then Head of Data Science, Mahana Mansfield. I would also like to thank Katharine Viner, who initiated and supported this project, and gave me access to the data, Helena Bengtsson, who helped design the analysis, Chay Woodford, who checked the results and created the graphs, Meghan McCarthy, who helped design the staff survey, and to all the Guardian staff who gave generously of their time. Annie Kelly coded the readers’ comments. Thanks also to Angela Phillips, Des Freedman, Mirca Madianou, and Lisa Blackman, for their advice and support, and to my reviewers for their invaluable feedback.
Notes
1. ”Below the line” refers to the comment thread; articles are sometimes referred to as being “Above the line.”
3. The US National Violence Against Women Survey estimated that 60% of “cyber-stalking victims” were women; Working to Halt Online Abuse found that 72.5% of the 3,000 reports of “cyber harassment” were from females; The Pew Research Center’s 2014 report found that women were more likely to experience sexual harassment and stalking (all cited in Citron Citation2014).
4. https://www.theguardian.com/community-standards (accessed 5 December 2017).
5. The data for 2016 was not for a full year, so the findings given here are for the years 2006–2015, unless otherwise stated. Only comments made on the Guardian’s website are included, not those left on Facebook or other social platforms.
6. This classification was done using genderize.io, which predicts gender based on first names. This is the best currently available approach (K. Wais Citation2016). The 746 names that could not be classified using this method were classified manually by the author, using the individual’s Guardian profile, or google search.
7. In a tiny minority of cases moderators either “watchlisted” (actively monitored) or pre-moderated a thread (if, for example, the author was seen as vulnerable).
8. Several Guardian journalists were subject to such abuse on social media.
9. Sent in February 2016.
10. Three people did not record their gender.
11. 41.5% of these were regular Guardian readers, i.e. they had visited the site at least eight times with a gap of no more than a week; 68% were from the UK or US.
12. The Women’s Media Center releases an annual report on the status of women in the US media: in 2017, they found that 62% of bylines across all media are male, while only 38% are female. A 2011 Women in Journalism survey of bylines in five national papers recorded a 78:22 split in favour of men.
13. Before 2007, ad hoc moderation would have been carried out by desk editors and the community team.
14. A “regular” writer is defined as someone who has written more than 100 times for Comment is free. Comment is free was the Guardian’s online opinion section.
15. It is possible that this is because Life & Style is a large and rather broad section, including a lot of disparate “light” subjects such as Family, Food and Drink, Health and Fitness etc.
16. There were also a number of criticisms of the methodology which were expressed in a polite or neutral tone (e.g. as a straightforward question) or which were valid: these were labelled neutral, not hostile.
17. Stephen Thrasher, a US-based, gay African American columnist.
18. Internal research in 2016 found that 69% of those who have posted comments are male, compared to 31% who are female, a commenter profile in line with that of other news sites (see Martin Citation2016; Pierson Citation2015).
19. For clarity, I do not mean that journalists should engage with abusive speech, which can cause real harm.
20. See her article, “A Mission for Journalism in a time of crisis” (The Guardian, 16 November 2017) https://www.theguardian.com/news/2017/nov/16/a-mission-for-journalism-in-a-time-of-crisis