New research sheds light on the behaviour of commenters on NYT website

On The New York Times website, people are more likely to comment on stories on weekdays rather than weekends, and the frequency of their contributions is also influenced by whether or not they interact with moderators or fellow readers on the platform, found a report published today by the Engaging News Project.

The research team analysed 9,616,211 comments posted on The New York Times website between 30 October 2007, when the NYT started allowing comments on news stories, and 13 August 2013.

During this period, the title redesigned its comment section in 2011, giving readers the ability to share their views on a story without having to leave the article page.

The report also found the number of comments increased in the year following the redesign, from 134,229 comments posted in February 2011 to 239,512 in February 2012.

Readers were also more likely to return to the site and comment following interactions with a moderator or NYT staffer, as well as with their fellow commenters.

This was also the case if an individual had one of their comments selected as a 'NYT Pick' or if their comment had been recommended by their peers – a person's activity increased, on average, from 0.2 comments in the month prior to receiving their first recommendation to 2.1 in the following month.

Other findings from the Engaging News Project's research highlighted the number of comments was higher on weekdays compared to weekends, but people were more likely to use uncivil terms and inappropriate language on weekends.

A question of moderation

"Bassey [Etim, community editor at The New York Times], who runs the moderation group at NYT, suggested that might be the case because there are fewer new news stories over the weekend, or there might be more opinion articles and magazine-style features posted then, which affects the use of negative comments," Dr Ashley Muddiman, research associate for the Engaging News Project, told Journalism.co.uk.

"It could also be that people are at work during the week, so they might keep up more with the news whereas on the weekend they might not necessarily be on their phone or their computer."

The length of a comment and the use of profane words also affected moderators' decisions to approve or reject a comment. Sentences containing words such as 'BS', 'damn' or 'hell' were 36.2 per cent more likely to be excluded.

"Sometimes we might look at the content and we might think something is inappropriate, but it's a little bit less clear. Name-calling a politician might not as strongly predict a rejection from the comment section, for example," Muddiman said.

"That's an interesting finding because many journalists and moderators in comment sections don't want to make subjective decisions about whether something is inappropriate or not, but with a profanity, it's very easy to say that it's clearly against the rules of the section."

'Meet our top commenters'

At The New York Times, moderation is mostly done in-house, with an additional team of external contractors working mainly on the overnight shift.

NYT also has a group of 'verified commenters', "a few hundred people" who can comment on the site without approval from a moderator, Etim told Journalism.co.uk via email.

The selection process to become a verified commenter is algorithmic, not manual, and takes into account elements such as the person's comment history and the ratio between their approved and rejected comments.

In November 2015, the outlet published 'Meet our top commenters', a feature looking at some of the people who often engage in conversation in the NYT's comment section and why they tend to get involved.

Greg Barber, director of digital news projects at the Washington Post and head of strategy and partnerships at The Coral Project, recently told Journalism.co.uk the Post has frequent back and forth conversations with its readers about what they expect from online comment spaces and that many news outlets could benefit from improving this experience.

"It's unclear what the organisation is looking to get from the reader, and in many cases there isn't any direct participation from the publisher or any process for highlighting good contributions," Barber said.

"There isn't that kind of value that could be if people knew someone was paying attention."

Update: An earlier version of this article said the average number of comments a person posted after receiving their first recommendation had increased from 0.2 to 2.1 per day. The number actually reflects their behaviour across 30 days.

Free daily newsletter

If you like our news and feature articles, you can sign up to receive our free daily (Mon-Fri) email newsletter (mobile friendly).