78

The question by Joel on Logo got me thinking. This has been raised on UserVoice, but has been declined.

Would someone with lesser rep or a less famous name have received the same number of votes/answers as Joel? I know the question was briefly closed, but it got reopened when the intent was made public. I think someone did try it, but I don't think their question was suitable for the experiment and it was soon closed.

Similarly, do answers by Jon Skeet and others with > 20K rep garner more votes than equally correct votes by others less notable? Is this because of their name, their rep, or the fact that the answer is correct? I would hope it's the latter, but perhaps we ought to find out.

I wonder if the following experiment would yield any useful information: If Jon created a second account under a pseudonym and posted a few answers along side slightly "worse" answers credited to Jon so we can see which ones garner the most votes.

Another experiment (though this would require a change to SO) could be to hide the identity of the poster for a period (a day?) but leave the (approximate) rep score visible. This might show whether people were using the rep as a guide. On a personal note, if there are two equally correct answers I tend to award my up-vote to the one with the lowest rep score - to spread "the love" around (so to speak).

NOTE

I'm not saying that questions and answers should be permanently anonymous, or that there shouldn't be a way of finding out the identity of a poster straight away. I just feel that the question or answer should be judged on its own merits rather than the perceived merit of the poster.

  • This is tagged as status-declined, but I recently saw some posts changed to anon. And I don't mean that the user changed their display name to "anon": there was no link to ta user profile. So perhaps this has been implemented, but can be used by staff/CMs/moderators only in emergencies? – Raedwald Oct 9 at 10:20
  • @Raedwald posts (used to?) get marked as being by "anon" when the account was deleted. I can't remember whether that's still the case or whether it's now "user<nnnn>". Posts can also be disassociated from accounts and the same would happen. What I was asking for here was for the posts to be still associated to an account, but just not immediate display the user name. – ChrisF Oct 9 at 10:28
  • 1
    @Raedwald They are marked as "anon" when a user later requests to be dissociated from the post under the CC license. – Sonic the Anonymous Hedgehog Oct 9 at 22:50

11 Answers 11

39

Great idea -- should be an option for both poster and reader.

Poster/answerer should be able to remain anonymous as an option, and change that option.

Reader should be able to set an option to, by default, not show names of posters/answerers. Maybe a global default setting, with mouse-over to reveal specific names when they really want to see.

  • 1
    I never thought about the poster requesting the anonymity - interesting. – ChrisF Jun 30 '09 at 21:01
  • 2
    I think this might be the solution I'm looking for. – ChrisF Jul 2 '09 at 20:14
  • as long as this anonymity expires in a day anyway, right? – forget it Nov 25 '09 at 10:25
  • At first, I thought a 5 minute delay would be appropriate as I state in my post below. However, upon additional reflection, a 10-15 minute window of opportunity, or until 3 answers have been posted - whichever occurs first - would be plenty of time. – IAbstract Feb 11 '10 at 16:43
  • 2
    This option should be set by the poster. Names and Reps will appear only if an answer is accepted. This would encourage answer acceptation. – Larry Apr 8 '11 at 8:17
  • I disagree it should be an option for new posts. I believe it should only become optional for established posts, where the impact of user-based voting has largely diminished. New questions swing wildly at low vote counts, biased users use this opportunistically. It's an unnecessary temptation. – Vogon Poet Oct 8 at 22:29
37

Maybe rather than hide the poster information for a day or two, you could just make the "answered by" an expandable area closed by default. This would "hide" the poster from people unless they explicitly clicked the link/button to show the answerer.

  • 2
    This is an interesting idea & one I (obviously) hadn't thought of. Could it be done via a roll-over rather than click? – ChrisF Jun 28 '09 at 18:35
  • Doesn't this defeat the purpose of anonymous? – Evorlor Mar 16 '15 at 22:57
  • 3
    You could also track how often people expand this section to gauge if they care who the poster was – Des Horsley Apr 30 '15 at 5:36
30

I just want anonymity as a poster because there are some very niche specific questions that I would much rather not have my name associated with.

  • 18
    Yes, that is how I found this thread. I want to ask a question, but I don't want my present or potential future employers to see it. It should still be associated with my account, but nobody should know I asked it but me. – Apreche Oct 19 '10 at 12:40
  • And at the same time all the rep points etc from that anonymous post connected to an account could still be accumulated anonymously.. – Irfan Jan 22 '12 at 16:54
  • That's a valid use-case, Quora has it already implemented. For SE sites, I do use a second account, where I put the questionable-questions, just to prevent problems like this. Also, I am aware about many very good programmers that do not have accounts on SO just because they are afraid of having a low or bad reputation. – sorin May 6 '12 at 14:09
8

The best way to sum this up is too look at john resig's all-time points to questions answered here. NB There are probably better examples of this on SO.

Answers (and comments) should be anonymous for a period of time to avoid the personality based voting. This would also help if we had mandatory downvote comments to avoid chasing the downvoter for retaliation.

8

The premise of the question is (imho) incorrect.

Answers from high rep users don't get more votes because they have high rep.

Certain users have high rep because they know how to answer questions effectively and quickly.

That being said, there is also a factor of who is answering the question. Users have reputations way beyond the number. If Jon Skeet answers my C# question, there's a pretty darn good chance that he's right. He's earnt that level of respect. That's a good thing for a question asker to have rather than hiding it on the grounds that doing so somehow creates a level playing field. That just means all answers are equally untrustworthy. Not good.

That doesn't mean Jon automaticaly "wins" any answer "contests" though. Other answers can be better, quicker, more to the point and so on and be rewarded as such. I've certainly seen this happen.

Good, timely answers will get upvoted and it doesn't matter who they come from.

  • 11
    "Answers from high rep users don't get more votes because they have high rep." Sure they do. People see the name and they upvote. – Evan Carroll Feb 11 '10 at 18:49
  • 4
    I agree with the points cletus made, except for one: I often found high rep users getting more upvotes even though there were better and earlier answers. – mafu Mar 23 '10 at 22:24
  • "Other answers can be better, quicker, more to the point and so on and be rewarded as such. I've certainly seen this happen." An example would help convince. – Carl Brannen Feb 25 '11 at 6:19
  • 2
    I would be interested in seeing some scientific tests that would validate or not these. I don't think that our personal oppinion would be relevant here. Let's see some statistics. – sorin May 6 '12 at 14:12
4

I don't understand why the author's answer must immediately show the rep, etc. See this question on SO. This is indicative of the higher rep author getting the up-votes and accepted answer when clearly another answer is more complete and provides a code example. Give a 5 minute window before the author's information is visible.

I would also add that this is not a singular, or rare occurrence. There have been a couple of occasions that this has happened to me.

Although this question on SO does not directly reflect a need for anonymity during a grace period, it does show a disparity in the amount of reputation one answer will get over another even though the answers are essentially the same. I view both answers as identical

In addition:

  1. Can I "inline" a variable... - although, props to the author of the accepted answer for editing his answer and referring readers to a better answer in the same thread.
  • 1
    In your example I see a 20k+ user with no upvote. Doesn't really prove the point. It was more a case of "first correct answer got upvote", then "upvoted answer calls for more votes", in my opinion. – Gnoupi Feb 9 '10 at 7:12
  • The times are not descriptive enough to provide a more detailed depiction of events. The top 2 answers appeared within a minute of each other; the answer wasn't selected until several minutes later. I had been writing an answer when the 1st answer appeared. Soon after (*within ~30 seconds), the 2nd answer appeared. I felt that the lower rep user, providing a higher quality answer, was/is given a disproportionate amount of reward for the answer. Even now, people continue to up-vote for the accepted answer (currently 2-1). – IAbstract Feb 9 '10 at 18:12
3

I have no doubt that upvoting and downvoting of questions as well as answers are biased to some extent by the reputations of the posters and the answerers.

I think all posters and answerers should be anonymous by default. The anonymity can be removed once a reader commits to a vote.

I would go further and suggest that if there is a way for the anonymity to be removed, and a reader chooses that option, they should be barred from voting on that question/answer.

This will ensure that all votes are based purely on the merits of the question/answer and not on the reputations of the posters/answerers.

I haven't thought about how this will impact the policy of reversal of votes.

2

I agree that it'd be nice to address the herd mentality type of voting. Sounds like a good experiment.

From a user perspective, however, I think hiding that information would be frustrating. If you're in a rush, sometimes you'd like to skip the white noise and go straight for the solution. Some people just post good answers the majority of the time.

  • 3
    I'm not denying that some people are usually correct and relevant with what they post. I suppose what I'm saying is that someone whose not famous (either generally or just within SO) posts an equally correct and relevant answer they aren't rewarded to the same degree as someone who is. – ChrisF Jun 28 '09 at 18:33
1

For what it's worth, I was the first to answer that Logo question, and I didn't notice (at first) who had asked it. (I noticed immediately afterwards, hence the jocular comment.)

If anything, I'd have been less inclined to answer the question if I knew it was from someone "famous". I suppose that's an argument for anonymity, from the other end.

1

I'm all for having complete, un-expiring anonymity when you want it. When asking a question, and when posting a comment. You should still be awarded points if you're up-voted too. If anybody abused the anonymity, there's always the flag - and moderators do/should have the power to look beyond that veil of anonymity.

  • I never thought about the poster requesting the anonymity - interesting. – ChrisF Jun 30 '09 at 21:02
  • 1
    In the physics den where I hang out, there are certain questioners who get minuses not because of the question they've asked, but instead because of their known beliefs (which knowledge is used as extra information about their question). I have no doubt that this question: physics.stackexchange.com/questions/5453 for example, would have gotten decent answers if it was from a beginner. Instead it got +2-9 and was closed as subjective and argumentative. But the arguments occur only in the comments. Are comments enough to justify argumentative? – Carl Brannen Feb 25 '11 at 6:29
-1

This is absolutely a great idea, for both Questions and Answers. User information is not in any way useful in generating quality questions or responses or deciding the appropriateness of those questions and responses. It invites bandwagoning and prejudice. This feature would greatly improve the quality of our metrics and greatly curb many bias-related challenges assailing our moderators..

Users on this site are here to make quality control decisions, and none of those decisions pertaining to the quality of a question or answer involve reputation or avatar or username or badges. Below is a complete list of the proper factors we should consider in our QC efforts:

Questions:

Upvote = "This question shows research effort; it is useful and clear."
Downvote = "This question does not show any research effort; it is unclear or not useful"

- Flagging reasons: - Spam: Exists only to promote a product or service, does not disclose the author's affiliation.

  • rude or abusive: A reasonable person would find this content inappropriate for respectful discourse.

  • should be closed...: This question is completely unclear, incomplete, overly-broad, primarily opinion-based or is not about the software that powers the Stack Exchange network as described in the help center, and it is unlikely to be fixed via editing.

  • a duplicate...: This question has been asked before and already has an answer.

  • in need of moderator intervention: A problem not listed above that requires action by a moderator. Be specific and detailed!

Answers:

Upvote = "This answer is useful"
Downvote = "This answer is not useful"

- Flagging reasons:

  • rude or abusive: A reasonable person would find this content inappropriate for respectful discourse.

  • not an answer: This was posted as an answer, but it does not attempt to answer the question. It should possibly be an edit, a comment, another question, or deleted altogether.

Note the clearly missing user name, reputation, and badges? Those things are only noise in the voting process.

More precisely, this should apply to new posts.

I believe established posts should afford users an opportunity to "brag" a bit. All new posts should be anonymous - questions and answers - however after it has become established the user info should be proudly displayed.

What is established? That would be a matter for discussion. I imagine criteria include:

  1. How many users have viewed the post?

  2. How many votes has the post received?

  3. How much time has passed since posting?

  4. Has the post been accepted by the OP (for answers)?

Handling comments: Certainly comments want to address the OP. In this case, anonymous posts receive a generic user/bot icon until it is established.

Fuzzy user display: A vague reputation gauge may be useful which still does not disclose the poster identity. But a reputation scale of 1-5 may make voters more comfortable. I still don't advocate this until someone makes clear how the "reputation" metric contributes to vote quality.

You must log in to answer this question.

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