18
$\begingroup$

Our community-modifiable default off-topic close reasons are

  • Self-study questions (including textbook exercises, old exam papers, and homework) that seek to understand the concepts are welcome, but those that demand a solution need to indicate clearly at what step help or advice are needed. For help writing a good self-study question, please visit the meta pages.

  • This question appears to be off-topic because it is not about statistics, machine learning, data analysis, data mining, or data visualization.

  • Questions about obtaining specific datasets are off-topic on Cross Validated. You are welcome to post this question to the Open Data site instead, thank you!

("This question belongs on another site in the Stack Exchange network" cannot be modified and will always be an option.)

I would like to add another default reason that in a friendly, polite, and helpful fashion says something like

This question appears to focus exclusively on programming, debugging, or performing routine operations within an application. It cannot be migrated to StackOverflow because it does not meet their standards or needs more information. You might be able to get an answer at one of the support links we maintain.

If you favor this, please suggest a better wording (and/or upvote suggestions you like); and otherwise, please explain why making this a default close reason might not be a good idea.

Important Edits

In attempting to implement this, I discovered a severe length limitation and pared back the proposed text (including the link) to fit the space provided. I managed partially to incorporate a suggestion made in the answer by Gung--there is a link to SO guidance for creating a Minimal Complete Verifiable Example.

Having succeeded in meeting this limit, I then (re-)discovered a limitation on the number of close reasons: we cannot have more than three! I am therefore asking for your suggestions about how to incorporate this suggested fourth reason within the three we are allowed. It probably needs to be rolled into the second close reason ("...not about statistics etc."), perhaps in this abbreviated form (which uses every allowable character):

This question appears to be off-topic because--to be understood and answered--it does not require any knowledge of statistics, machine learning, data analysis, data mining, or data visualization. If it focuses on programming, debugging, or performing routine operations within a statistical computing platform, visit the support links we maintain.

Previous discussions on this topic occurred at

$\endgroup$
3
  • $\begingroup$ Yes, this is a very good idea that is coherent with my recent question on Meta: meta.stats.stackexchange.com/questions/2418/… $\endgroup$
    – Tim
    Commented Mar 6, 2015 at 17:18
  • $\begingroup$ @Tim Thank you. I didn't find your question in my search because it does not have the close-reasons tag. $\endgroup$
    – whuber Mod
    Commented Mar 6, 2015 at 17:47
  • $\begingroup$ I didn't find this tag while searching for it either... $\endgroup$
    – Tim
    Commented Mar 6, 2015 at 18:05

2 Answers 2

Reset to default
16
$\begingroup$

I like the idea, and I like the current phrasing. I think we'll end up with a system that operates more smoothly with this as a default option. I think having to type a custom comment is sometimes discouraging for people, and so they either go with migration (when it will just have to be closed on SO), or the "not about statistics..." option (which often isn't quite true).

I do worry about making the suggested text longer, but it would be nice if it could be more specific than "...it does not appear to meet their standards." I wonder if something like "...it does not appear to meet their standards (e.g., coding questions need a MCVE reproducible example)." I suggest this because I would guess 99% of the threads that don't meet their standards do not by virtue of lacking a MCVE.

Update:
Hmmm, I would prefer they be separate, but if you can only have three then we're stuck. Still, I would prefer the reasons be more distinct. The current text makes it possible that someone would misread it and slur the two reasons together. ('What do you mean that my programming question doesn't require any knowledge of statistics to answer?') I might try:

This question appears to be off-topic because EITHER it is not about statistics, machine learning, data analysis, data mining, or data visualization, OR it focuses on programming, debugging, or performing routine operations within a statistical computing platform. If the latter, you could try the support links we maintain.

(The custom reason comment allows for 500 characters. If this reason must be 100 less, we get 400. My suggested text above is 377 characters.)

I'd still like to get the reproducible example in. We could also try:

This question appears to be off-topic because EITHER it is not about statistics, machine learning, data analysis, data mining, or data visualization, OR it focuses on programming, debugging, or routine operations in software without a reproducible example. If the latter, you could try at the links here.

That clearly required me to use very sparse text, but it is 395 characters.

$\endgroup$
8
  • 5
    $\begingroup$ +1, but I would replace "MCVE" by "reproducible example" (while still keeping the link). An abbreviation will only be confusing. $\endgroup$
    – amoeba
    Commented Mar 6, 2015 at 9:57
  • 1
    $\begingroup$ +1, also agree with @amoeba $\endgroup$
    – Andy
    Commented Mar 6, 2015 at 11:04
  • 1
    $\begingroup$ +1 I'd go with "small reproducible example", myself (I don't care if it has 3 numbers or 12, as long as I can easily copypaste) $\endgroup$
    – Glen_b
    Commented Mar 7, 2015 at 7:33
  • 1
    $\begingroup$ Thank you, Gung. I tried to implement this suggestion, but was stymied by system-imposed limits. A fuller explanation appears in edits to the question. $\endgroup$
    – whuber Mod
    Commented Mar 7, 2015 at 18:24
  • $\begingroup$ @whuber, is it that you have a character limit? $\endgroup$
    – gung - Reinstate Monica
    Commented Mar 7, 2015 at 18:37
  • $\begingroup$ @gung Yes--and it is 100 characters less than you are allowed when writing your own custom close reason ("Other (add a comment explaining what is wrong)"). $\endgroup$
    – whuber Mod
    Commented Mar 7, 2015 at 18:38
  • 1
    $\begingroup$ @whuber, I updated my answer. See what you think. If we are allowed 400 characters, it would just make it. $\endgroup$
    – gung - Reinstate Monica
    Commented Mar 8, 2015 at 16:31
  • 1
    $\begingroup$ Thank you: that's getting to the crux of the matter and addresses the issue that was troubling me most. $\endgroup$
    – whuber Mod
    Commented Mar 8, 2015 at 17:22
7
$\begingroup$

The suggestion makes sense but I have yet to make up my mind regarding whether it would be better to have an extra field for this or to amend the option

  • This question appears to be off-topic because it is not about statistics, machine learning, data analysis, data mining, or data visualization.

to

  • This question appears to be off-topic because it is not about statistics, machine learning, data analysis, data mining, or data visualization, or it focuses too heavily on aspects of programming/debugging. For this you might be able to get your question answered at one of the support links we maintain.

Probably not the best wording for it but it might be a starting point for an alternative suggestion.

As an additional comment I wonder whether we should migrate such questions at all. There is of course no doubt regarding ones in the spirit of

  • "How do I do this thing in R/SPSS/Matlab/Stata"

What I am thinking of are questions of the sort

  • "I'm trying to program this estimator but I don't know where the error in my code is"

because to figure this out it might require the statistical knowledge behind the given estimator or test statistic. I think this type of question would be better served here given that many of our users have extensive knowledge of statistical packages but rarely have the chance to apply this knowledge here in the statistics context. If anything this skill is mostly used for illustrating the theory behind something. But statistics shouldn't be just about the theory (only my view) and an important part is the application for which this kind of question would be useful. Maybe we are restricting ourselves too much by migrating too often.

$\endgroup$
4
  • 3
    $\begingroup$ These are reasonable points. It seems to me there are 2 parts to your answer here. 1 pertains to combining the two closing rationales. 2 raises the issue of whether all coding Qs need migration. Your 2 doesn't seem to be quite an answer to this meta Q, but it is a good Q. Coding migration has been raised & debated over & over on meta.CV--w/o much progress :-(, but that specific issue is a novel & worthwhile distinction. It may do well as another meta.CV thread. $\endgroup$
    – gung - Reinstate Monica
    Commented Mar 6, 2015 at 4:26
  • 3
    $\begingroup$ Re 1, I prefer the two closing reasons to remain separate. I would further prefer if the "not about stats..." were used less often (I think it strictly applies infrequently). To my mind the 2 rationales are distinct & having them as differing options would be best. $\endgroup$
    – gung - Reinstate Monica
    Commented Mar 6, 2015 at 4:28
  • 1
    $\begingroup$ You're right, my second point is off-topic. It's an issue I hold dear and I saw a relation to whuber's question :) It makes sense to have the two separate, especially with the text you added to the option in your answer. $\endgroup$
    – Andy
    Commented Mar 6, 2015 at 11:06
  • 2
    $\begingroup$ You were prescient, Andy: limitations in the length of the text and in the number of default reasons compel us to adopt your approach, using something very much like your wording. Please don't hesitate to comment further on the edits I have made to the question. $\endgroup$
    – whuber Mod
    Commented Mar 7, 2015 at 18:24

You must log in to answer this question.

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