2

In a short discussion between myself and moderator ff524 about a deleted question. Moderator ff524 mentioned,

There are other sites that are not as strict about moderation - Reddit, etc. SE is deliberately different in its moderation policy, because we try to be a home for people who like answering high quality, on topic questions. But some of our users enjoy contributing to those other sites, too ;)

Emphasis mine.

I agree with Moderator ff524 on all points. But this led me to think about the effects of Anonymity on both questions and answers in Academia specifically.

In this particular case, the author of the deleted question has some experience and interaction with SE sites, specifically Mathematics SE. But purely on mathematics topics.

With the following variables on anonymity in play (not exactly rigorous, but will suffice in the meantime):

Anonymity:

  • Binary: has an equivalent real-life ID associated, or a pseudonym.
  • Effect: if real-life, then words said would carry over to their IRL identity. If pseudonym, then there is a degree of buffer between the two.

Reputation:

  • Integer: based upon a voting system reflected up on a user's contribution.
  • Effect: more privileges and respect afforded to the user and can be used as a rough yardstick to measure the experience of the user to the SE model.

What is the effect of Anonymity on both the quality of questions and answers?

The various logical derivatives:

  • Does real-ID mean higher-quality questions?
  • Does real-ID mean higher-quality answers?
  • Does pseudonyms mean higher-quality questions?
  • Does pseudonyms mean higher-quality answers?
  • Does real-ID mean lower-quality questions?
  • Does real-ID mean lower-quality answers?
  • Does pseudonyms mean lower-quality questions?
  • Does pseudonyms mean lower-quality answers?

In my experience at the Workplace as well as Academia, I would presume that Anonymity has a role to play, but not to the extent that I theorize.

There are great question and answer contributions from users from both sites whether with a pseudonym or real-ID.

But in general, poor-quality questions and answers are from pseudonyms; whether due to inexperience or the buffer (between a person's actions and his/her real-life identity) offered by a pseudonym.

I could be entirely wrong after all, would someone with more expertise on the matter care to comment?

  • Mm... dunno. I write on the net both with my real name, here and on other SE sites, and with a pseudonym in an Italian technical forum, where I usually answer questions about electronics, physics, metrology and mathematics (now very rarely due to a chronic lack of time). I try to do my best to give good, possibly useful answers under both identities, and probably my better answers are really the technical ones, given under pseudonym (which, in some sense, could be a pity). – Massimo Ortolano♦ Nov 13 '17 at 21:36
  • the deleted question link is broken, FYI – Federico Nov 14 '17 at 12:28
  • You can't see it because the question was deleted, – Frank FYC Nov 14 '17 at 18:43
6

If you take a look at the top page of users, the split is slightly in favor of pseudonyms over real names (4:3 advantage, roughly).

I'd argue that the quality of answers is independent of the use of a pseudonym. The quality of the questions depends on experience with SE sites rather than just being "experts" or not. We've had many excellent "signed" and "anonymous" questions. However, as you suggest, it's unlikely someone will "sign" a bad question with their real name.

  • 1
    I can recall a number of cases of mod worthy bad behavior by users who have provided their real names. – StrongBad♦ Nov 13 '17 at 22:08
  • Several of those who use a pseudonym, however, have a link to their home pages from which the real name can be inferred. I think the comparison should be done between totally anonymous and others. – Massimo Ortolano♦ Nov 13 '17 at 22:09
  • @MassimoOrtolano I tried to do that and only found 7/36 without much information in their profiles. – StrongBad♦ Nov 13 '17 at 22:13
2

Looking at our first page of users 7/36 users provide limited information in their user name or profile (i.e., are anonymous). Three of these users are, or were, moderators. At least one of these users (me) was not always anonymous. On a random page of users 15/37 are anonymous despite having reputations between 2831 and 3365. These differences could be related to the quality of contributions or an indication that higher rep users are more likely to not be anonymous.

We can also look at the average vote counts for questions and answers by anonymous and non-anonymous users. This is sounding like a question for the data explorer, but taking a very small sample of the top three non-anonymous and anonymous users reveals that the top three non-anonymous users have a combined 3628 answers and a combined rep of 352158, for an average rep of 97 per answer (I ignored the number of questions). The top three anonymous users have 2083 answers for a combined rep of 277104, for an average rep of 133 per answer. This suggests that anonymous users may provide better answers.

In summary, as a biased anonymous user, it looks like anonymous users provided better answers :) As an academic who likes to butcher statistics, I bet my findings will not replicate.

  • But, for instance, the top one, who is not anonymous, hasn't answered any question in recent times, when many of our questions reached the HNQ list with its boost effect on reputation ;-) – Massimo Ortolano♦ Nov 13 '17 at 22:21
  • @MassimoOrtolano the non-anonymous users would need an additional 653 days of rep cap performance to close the gap assuming they managed to do this without answering any more questions. – StrongBad♦ Nov 13 '17 at 22:31

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .