3
$\begingroup$

I have come across a vibrant question - how does one visualize results of post hoc Nemenyi tests - on Crossvalidated. This question is often raised in my institute and by students or peers at conferences. Most times people waste hours to code self-made solutions or draw it by hand in a graphical program. I was angry enough that I hacked a solution to it. But the has been marked as off-topic on Cross Validated, and I miss reputation to do anything with it.

What should I do with it? The question is valid, perhaps it should go to Stack Overflow, but then - why wasn't it moved instead being tagged off-topic? How do I report it to be moved if it is closed as off-topic and I don't see any flag options? I can only edit the question but I don't think that is needed at all.

$\endgroup$
1
  • 5
    $\begingroup$ Briefly, "how does one visualize results of post hoc Nemenyi tests" is not really an accurate summary of your original question, as the question is entirely about how to code a plot in specific software, which was why it was closed. (Disclosure: I voted to close, and stick by that view.) But that question would be entirely acceptable, indeed excellent, as a question here. $\endgroup$
    – Nick Cox
    Nov 19, 2014 at 10:41

3 Answers 3

4
$\begingroup$

The question is from January - much older than the 60 days in which a question can be migrated.

It was most likely not migrated because not enough closers reached the migration consensus (3 of the 5 people closing need to have selected Stack Overflow as a migration target for that to have happened).

As it stands - this is not a suitable question for Stack Overflow, so a migration, if it had occurred would have most likely been rejected.

$\endgroup$
3
$\begingroup$

I sympathize with your predicament. After all, it's not easy being green. As @Oded states, the question could not be migrated, even if it were re-opened, because it is too old. I do think a similar question could be on-topic on Stack Overflow. Why not ask an analogous question on SO yourself, and then provide your own answer? Asking and answering your own question is unusual, but "Stack Exchange has always explicitly encouraged users to answer their own questions".

$\endgroup$
3
$\begingroup$

I sympathize with your request, because I both edited and upvoted the question on January.

You could suggest an edit to let it be on topic. Here is one suggestion:

Currently, it is written like the following:

Could you please tell me which R function/library (not necessarily R) to use to plot Nemenyi's test result with the critical differences?

Let it be on topic:

What is the correct way to interpret a Nemenyi's test plot? Also, is there any R function/library (not necessarily R) to use to plot Nemenyi's test result with (and without) the critical differences?

It would require two reviewers to approve the edit suggestion. The edit would not conflict with the author's intent, because it would only have information added. After the edit being approved*, moderators and 3k rep users could reopen the question.

Then, you could enlighten the community with an answer split in:

  1. The theoretical explanation about what is and how to interpret the Nemenyi's test plot.
  2. A programming solution to produce such graph.

*it is not guaranteed it will be approved.

$\endgroup$
2
  • 1
    $\begingroup$ +1 - I believe a lot of the questions that are migrated would take minimal effort to edit to be on topic. Typically they are of the form I need code to do X - although X is sometimes ambiguous and could benefit from discussion about the statistics here as opposed to just rote applying a set of code. (That isn't a criticism of the way things are currently done though - I think the balance of migration at this time is working out reasonably well given there will always be ambiguity.) $\endgroup$
    – Andy W
    Nov 20, 2014 at 12:39
  • 3
    $\begingroup$ @AndyW: That's true, but it's less obvious that it would help whoever asked the question. It's not a punishment to have a question migrated to SO. Asking your own statistical question inspired by someone else's programming one might be better. $\endgroup$ Nov 20, 2014 at 15:11

You must log in to answer this question.

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