16 Comments

Summary:

Data from Disqus, which offers a comment-hosting service for websites, seems to show that the use of pseudonyms not only produces more comments, but also comments of higher quality. As interesting as the data is, however, it’s unlikely to settle the ongoing debate over online identity.

282428943_322a2027b4_z

The battle over online identity has been going on for almost as long as the internet has been around. Should users — including members of social networks or commenters on blogs and other websites — be forced to use their real names, or allowed to remain anonymous? Do pseudonyms allow for more open discussion, or encourage trolls and flame-wars? Disqus, which offers a comment-hosting service for websites, came out with some data from its network that seems to show the use of pseudonyms not only produces more comments, but that the comments generated are also of higher quality. But as interesting as the data is, it’s unlikely to settle the debate over how to handle online identity.

Disqus says that it drew the data for its infographic (part of which is embedded below) from the comments on more than one million websites that use its hosted service, a network that the company says accounts for more than 600 million monthly visitors and 60 million comments commenters. While that’s obviously only a small proportion of the overall web, it’s still a fairly large sample — and Disqus says the data shows that “the most important contributors to online communities are those using pseudonyms,” since they contributed over 60 percent of the comments in the sample. Those using pseudonyms were also more active than other commenters, Disqus said, posting more than six times as frequently.

Pseudonymous users made more — and better — comments

As some critics of the report noted on the Hacker News discussion forum, more comments don’t necessarily make for a great community, especially if they are primarily trolls or attempts at flame-bait. But it wasn’t just frequency or the sheer number of comments that stood out, according to Disqus. The company also tried to assess the quality of those comments by looking at how many were rated highly by other users (i.e. “liked”) and how many drew replies, as well as how many were flagged as being offensive, how many were marked as spam and how many were ultimately deleted by the site.

Although many critics of pseudonymous or anonymous comments — including those who have turned off comments on their blogs — suggest that a lack of real names leads to an overwhelming amount of hateful and offensive comments, the Disqus data seems to show that this isn’t really the case. According to the company’s quality rankings, more than 60 percent of comments using pseudonyms were positive, and almost 30 percent were rated as neutral, while only 11 percent were rated negative. In fact, the company says that a greater proportion of pseudonymous comments were positive than those that used real names (that is, logged in with Facebook or some other identity service).

This news was celebrated by some advocates of pseudonymous behavior online, including Christopher “Moot” Poole, the founder of the online community known as 4chan — a group of forums that have become notorious for their bad behavior. Despite that reputation, Poole has spoken a number of times (including at a TED conference in 2010) about the value of anonymity and pseudonymity when it comes to online culture, and how requiring real names is not the right response to negative or offensive speech.

Perhaps a lack of real names doesn’t mean chaos after all

Others (including us) have also argued for a more flexible approach to identity, particularly during the launch of Google+, which required real names from users, and sparked a wave of criticism from those who are routinely known by their online pseudonyms (as some Google executives are, including Google+ lead Vic Gundotra). Google later agreed to modify its policy, although the details about how that is going to work haven’t been fully revealed. Facebook, meanwhile, continues to require real names, although it has made exceptions for prominent users such as Salman Rushdie.

It’s difficult to draw a hard-and-fast conclusion from the Disqus numbers, in part because it isn’t clear whether the distinction between pseudonyms and real names is reliable or not. As commenters at Hacker News pointed out, the company seems to have categorized everyone who used a Facebook login as a real name, and everyone who didn’t as a pseudonym — but some users who don’t log in with their Facebook identity are likely also their using real names, even if they haven’t been verified by a third party (Disqus founder and CEO Daniel Ha said the company plans to expand on the data in a future blog post).

While the Disqus data may not be conclusive, however, it does seem to show that pseudonymous speech doesn’t necessarily lead to worse behavior than any other form of online identity, and may even lead to better behavior (at least for those who see user comments as being valuable at all). And that serves as a bit of an antidote for those who have been pushing a “real names are better” agenda, including Google and Facebook, and a boost for those — such as Twitter — who don’t seem to care what a user’s real name is, so long as they aren’t pretending to be someone famous.

Post and thumbnail photos courtesy of Flickr users Klobetime and

You’re subscribed! If you like, you can update your settings

  1. Mohit Khurana, CFA Tuesday, January 10, 2012

    It is nearly impossible to strike an exact relation between use of pseudonyms and the quality of comments…

    1. Yes, I agree Mohit — it is a difficult problem. Thanks for the comment, and for using what appears to be your real name :-)

      1. Hey Mathew, yes it is my real name :-)

    2. I agree. I have never used my “real” name on the internet (hearing in the early 90s that doing so would lead to ID theft), but I have been consistent in using the same pseudonym whenever possible.

      For about fifteen years I was very active on forum sites using the same pseudonym. To me, if has become just as valuable as my “real” name. When I post to CNN, The Deseret News or The Salt Lake Tribune, I get noticed and messages are often posted to me at a forum site about what I said.

      I think the problem is when individuals have so many “identities” and “pseudonyms” that any contributions, positive or negative, get lost. Who cares if one does not know who you are?

      Right now, I don’t even look at the name of a poster to a newspaper or forum unless the content is useful and often when the post is useful, it comes from someone I know, “real” name or not.

  2. What I find interesting is that the negative numbers don’t vary a great deal. Whether the quality of the comments is good or not generally doesn’t become an issue, but the “troll” aspect is always thrown out there. Seems to me that trolls are just about as likely regardless of whether someone uses their real name, a pseudonym or remains anonymous.

    1. I agree, Karen — that’s been my experience as well. I’ve seen some pretty horrible behavior from commenters using their real names. Thanks for the comment.

  3. One issue here is that the negative signals are only the posts that are flagged, spam, or deleted, while positive comments are the ones that users like. So the negative signals require active moderation, while positive are user input. The amount of moderation is going to heavily bias the data here. Also, negative comments are much more destructive per post than a comment that gets a like.

    I would also suspect that users using real identities are less likely to use the like button, hence more “neutral” comments (although this wouldn’t account for the entire difference).

    1. Those are fair points, Ryan — especially the one about negative comments requiring more moderation, and therefore more effort, than a positive one. Thanks.

  4. There is not just one answer to this question. The answer is highly contextualized and totally dependent on what is being commented on.

    If you have a domestic abuse or suicide support forum that is definitely a context for anonymity.

    Where as a site that is for professionals to share tips or information. This might cal for real identities.

    Quality is not really part of this. You are going to get high quality posts on the suicide support forums with anonymity and high quality posts on the professional site.

    I suppose the above are extreme examples but you get the point.

  5. Hey Mathew! Quick correction — the Disqus 60mm number is commenters, not comments.

    1. Ah, okay — my mistake, Daniel. I will fix that.

  6. Stephen Candelmo Tuesday, January 10, 2012

    I am not sure how identity equals quality and therefore I doubt that correlation can be made through counting “likes” or responses. As said before, it all depends on the context. Obviously, blogging comments are only one type of social content. There are certainly areas that real names impart a higher level of trust and therefore value. The world of online reviews, for example, is beginning to suffer from this problem more and more whether through anonymous or pseudonym based reviews masking disingenuous motives. A real name in that context I would surmise would be more valuable for the reader. In my opinion, multiple identities are counter-productive toward promoting social media’s future authenticity.

  7. I’m curious what’s going on with the chart…is the fact that bars representing 51% and 61% appear visually equal intentional?

  8. Do you people think there will be changes to fake name to real name in this internet world

  9. health online Thursday, March 8, 2012

    After going over a handful of the blog posts on your website, I honestly appreciate your technique of blogging. I book marked it to my bookmark webpage list and will be checking back in the near future. Please visit my website as well and tell me how you feel.

  10. Kiyan Ridley Monday, March 12, 2012

    see the first comment in the pile.. health online.. case closed. Disqus can work tho..

Comments have been disabled for this post