8
$\begingroup$

I approached a question the wrong way and I would like to ask it again from a different point of view, but it already has an answer.

I asked this question: Estimating quantiles with linear regression?

The one answer I received made me realize that I don't have a clear idea of prediction intervals. I should not have mentioned them in the question. However, the answer doesn't answer my question, i.e., how (if at all possible) to estimate quantiles from linear regression. The answer also mentions quantile regression and/or Bayesian inference. This helped me realize that for various reasons I would benefit from an answer within the boundaries of linear regression and frequentist inference, or at least that showed me why it doesn't make any sense to try to estimate quantiles within these boundaries.

Usually, I would delete the question and write a new, clearer one, adding all these details and removing mentions of prediction intervals. However:

  1. I cannot delete it since it has an answer (well, I could delete it, but I understand it's not good practice).
  2. Removing all the part on prediction intervals will make the only answer look meaningless, and this also is not nice towards to the answerer.

Any suggestions? Maybe write a new question and flag my same old question for duplicate? Would this preserve the reputation of the answerer?

$\endgroup$
12
$\begingroup$

It doesn't really sound like you are [would be] asking a duplicate question. I think a useful way to define a duplicate question isn't a question that uses some of the same terms, backstory, etc., but a question that would be answered by the answer in an existing thread. If that isn't the case, the new question shouldn't be considered a duplicate, in my opinion. Consider the boilerplate text I sometimes add as a comment when I vote to close a question as a duplicate (taken from here):

I think you will find the information you need in the linked thread. Please read it. If it isn't what you want / you still have a question afterwards, come back here & edit your question to state what you learned & what you still need to know. Then we can provide the information you need without just duplicating material elsewhere that already didn't help you.

If the information at the other thread is different from what you need, then it shouldn't be considered a duplicate.

Therefore, what you need to do is just make that sufficiently clear. For example, what you have written in your first three paragraphs makes what new / distinct information you want to know pretty clear to me.

$\endgroup$
  • 8
    $\begingroup$ I'd also add that it's often helpful to me to see a couple of sentences at the start of the new posting clarifying that you've previously asked a somewhat similar question (include a link) but now realize that you need to ask a different question. I think that gives pause to someone who searches and finds an apparently duplicate question -- by you no less -- so that they see you're not just one of those guys who asks the same question again and again. $\endgroup$ – Wayne Sep 15 '16 at 17:40

You must log in to answer this question.

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