4
$\begingroup$

I'd argue that this community are probably the best fit community to answer questions on R, Stata, SPSS, etc. but currently these are deemed "non-statistical questions" and closed.

These programs are pertinent tools for anyone doing serious statistics and navigating around them is an important part of statistics. You may argue they are the cornerstone of applied statistics.

I therefore ask why not allow statistical questions in which the user may be experiencing difficulties with their software.

$\endgroup$
4
  • 9
    $\begingroup$ One of the strongest arguments is that such questions can & are asked, & in many cases of the re-occuring questions already have been answered on StackOverflow.com, so it is pointless to duplicate it. $\endgroup$
    – Tim Mod
    Jan 28, 2020 at 9:24
  • 10
    $\begingroup$ Some variant of this is perhaps the most commonly recurring question on Meta but I'd argue that "statistical questions in which the user may be experiencing difficulties with their software" are already allowed here and frequent. The long-standing dividing line is whether the question requires statistical expertise to answer, in which case mentioning the software is germane. Otherwise the leading response is that given by @Tim that this is what Stack Overflow is for. $\endgroup$
    – Nick Cox
    Jan 28, 2020 at 10:26
  • 4
    $\begingroup$ I have some scope to comment specifically on Stata. Statalist has seen nearly 250,000 posts since relaunch as a web forum on 31 March 2014. Stack Overflow has seen about 3000 Stata questions since 10 October 2009. Even with a multiplicative factor to turn threads into posts and an addition of whatever threads were deleted. SO is way behind and Statalist is way ahead. No hostility or negativity there: I am active on SO too answering Stata questions. $\endgroup$
    – Nick Cox
    Jan 28, 2020 at 10:30
  • 1
    $\begingroup$ Naturally anybody with more experience can and should comments on R, SPSS, etc. It's my impression that de facto SO is the leading R forum, having superseded R-help some while back. $\endgroup$
    – Nick Cox
    Jan 28, 2020 at 10:37

1 Answer 1

Reset to default
11
$\begingroup$

This was discussed in a previous thread some time ago, but I think it is reasonable to bring it up again.

There is a continuum of questions about software. Some are strictly programming , some are somewhat statistical and some are very statistical and only use the software as an illustration.

The current policy, as I understand it, is that strictly programming questions are off topic, very statistical questions are on topic, and then ... there's a gray area.

As @Tim pointed out in a comment, many of the strictly programming questions (e.g. questions about merging files) are already answered on Stack Overflow. We don't need to duplicate answers and people on Stack Overflow may be more expert in this.

And questions that only use software as an illustration, or that ask us to interpret output that could come from any software, are clearly on-topic. E.g. if you are asking about interpreting a covariance matrix it doesn't really matter if it came from R, SAS, Python or whatever.

But I would welcome more guidance about the gray area. E.g. "I ran glm() in R and got these errors" - right now, a lot of these get migrated. But I think that, at least for some errors and some languages, we have experts here.

$\endgroup$
2
  • 3
    $\begingroup$ You touch on the grey area, but I don't think writing more guidelines would help. There are blatant newbie questions of the form "What software should I use?" that unfortunately are all too obvious. There are intricate questions with a heavy software component where often it is not formal rules that would decide for or against migration but detailed expertise in the software and the statistical topic, which should inform voting. $\endgroup$
    – Nick Cox
    Jan 28, 2020 at 10:34
  • 3
    $\begingroup$ Here's one rule of thumb: If the same question could easily if the OP were using different software, it is possibly a statistical question. It's fallible: how do I loop in X? is a programming question, regardless. $\endgroup$
    – Nick Cox
    Jan 28, 2020 at 14:56

You must log in to answer this question.

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