13
$\begingroup$

I came upon this question which I thought was an interesting probability question and a detailed answer so I upvoted both.

What are the mean and variance of a 0-censored multivariate normal?

Then I was puzzled noting it was asked and answered by the same person literally within a minute. In fact, given the amount of code and algebra in the answer, I wager the answer was written before the question.

I don't know if it's just for reputation or influencing the content of the site; but is it technically a violation of site etiquette to post personal work, albeit interesting, this way?

| |
$\endgroup$
  • 18
    $\begingroup$ No, it's definitely not a violation of site etiquette. See our help: stats.stackexchange.com/help/self-answer. (If you look closely when asking a question you can see the site explicitly encourages self-answers.) Related meta threads include stats.meta.stackexchange.com/questions/1844 and stats.meta.stackexchange.com/questions/2142. $\endgroup$ – whuber Feb 1 '18 at 21:48
  • 6
    $\begingroup$ Reputation is only a game! Content is surely subject to peer pressure, as always. $\endgroup$ – Nick Cox Feb 1 '18 at 21:56
  • 12
    $\begingroup$ It wasn't just "within a minute". Both Q and A were posted exactly at the same time (timestamps are identical up to the seconds), which is possible by design of Stack Exchange: one can write both on the same page and submit with one click. Just look closer at the "Ask Question" form. $\endgroup$ – amoeba Feb 1 '18 at 21:56
  • 7
    $\begingroup$ Thanks for the feedback. I see now how this can be a useful approach. Readers can offer different perspectives to your problem as well as critiquing your approach. This method may in fact be better than posting both in the same "question" and dually asking for alternate approaches and critiques... I have seen those kinds of questions go unnoticed. $\endgroup$ – AdamO Feb 1 '18 at 22:31
  • 5
    $\begingroup$ Self-answered questions have been engineered to be hard to game for reputation (if you have an interesting new question you can answer, try it some time and see). You can get reputation for a good question or answer (or both) by attracting votes or bounties from others, as with other questions and answers. That form of reputation gain is not a problem. $\endgroup$ – Glen_b Feb 2 '18 at 0:08
  • 2
    $\begingroup$ I hope, now this has been brought to light, more people add interesting questions & self-answers to help share their knowledge :) $\endgroup$ – Firebug Feb 2 '18 at 15:17
  • 7
    $\begingroup$ As the author and answerer of the question at hand: yes, I derived the answer / wrote the code before considering asking a question here (though of course I searched if the result was already on the site first!). I figured others in the future might want the answer to the question, since it's annoying to compute. I also got a suggestion from whuber that made me realize I could have done it less annoyingly, which will be nice to know in the future if a similar thing comes up. :) The 0.5% of my total reputation that it's gained me thus far was definitely not a factor. $\endgroup$ – djs Feb 2 '18 at 15:58
  • $\begingroup$ @Dougal Then your example almost perfectly illustrates why a poster would take such an approach. Thanks for showing me how to better use the site! $\endgroup$ – AdamO Feb 6 '18 at 21:44
  • $\begingroup$ @AdamO Well, in retrospect perhaps I should have asked before deriving the answer. Then whuber's suggestion could have saved me some time. On the other hand, someone else might have started working on it in parallel, which was what I was trying to avoid. $\endgroup$ – djs Feb 6 '18 at 21:45
27
$\begingroup$

There is nothing wrong about it. It is perfectly ok to answer your own question, the system even allows it and changes the name of "answer" button to "Answer your own question". This was done in the past by many users, even high-reputation users did start their own question and answered them, e.g. @amoeba in here. Usually you ask a question if you want to learn something, but there is nothing wrong with starting your question if you want to share your knowledge or solution of some problem. As in @amoeba's case, creating and answering question by yourself may be also a good way to produce answer for a generic question that appears on the site over and over again, so we have an answer for the reference.

| |
$\endgroup$
  • 2
    $\begingroup$ Also, (albeit not the case here) it's a good tool to provide content for beta sites. $\endgroup$ – Firebug Feb 2 '18 at 15:16

You must log in to answer this question.

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