7
$\begingroup$

This question was originally on stackoverflow and I flagged it to be migrated to this site. The question is framed in terms of choosing between two different SPSS procedures, which is a statistical question - not a programming one.

I flagged it for migration to here because it is a good (and subtle) question as well - what weights to use when you only have summary information. It is a question that has perplexed others as well, so I think it is appropriate to discuss on this site.

Temporary closure is reasonable to clarify details of what the two methods do (so people do not need to look up the details in alternative documents), but that is not why the question was closed by the 5 individuals.


To be a bit more general, while many questions come in the form of How do I choose between A and B in software Y they can always be broadened to other contexts by stating the differences in the underlying models - which is software agnostic.

Specifically contrasting two software procedures is just a practical way most people encounter such issues. It is a steep hill to climb for many users to be able to formalize different software procedures in equations.

$\endgroup$
  • $\begingroup$ I'm currently getting the post is locked so I cannot edit it (which I believe is necessary to prompt a re-open vote). $\endgroup$ – Andy W Mar 23 '16 at 13:02
  • 2
    $\begingroup$ Even on re-inspecting the thread, it primarily looks like a 'how do I use SPSS' question. If you can edit it to make it more overtly statistical / more software-neutral along the lines you discuss above, I would happily vote to reopen. I do see that the edit link is not working. Perhaps the moderators can fix that. $\endgroup$ – gung - Reinstate Monica Mar 23 '16 at 13:56
  • $\begingroup$ @gung Apparently the "Community" bot automatically locked this question because it had been migrated in to our site and then rejected. I have removed that lock. $\endgroup$ – whuber Mar 23 '16 at 14:15
  • 1
    $\begingroup$ I don't agree @gung. The poster stated is method 1 or method 2 correct after describing how to do each procedure. I agree that the nature of the procedures should be clarified (so closing for unclear and asking the op for clarification would be reasonable) - but that wasn't the reasoning given for its closure. $\endgroup$ – Andy W Mar 23 '16 at 15:15
  • $\begingroup$ "How to incorporate weight into linear regression in SPSS" seems to clearly frame an SPSS coding issue. Most such questions are closed. The underlying statistical issue is on topic and if it was framed to focus on that, it might go quite well. $\endgroup$ – Glen_b Mar 23 '16 at 22:30
  • 2
    $\begingroup$ @Glen_b - you have to read the whole question. After asking that the OP then goes on and describes two separate procedures and then asks which one is correct. (If you looked at the comments, I had already provided code in a prior question by the OP.) We help people better formulate questions all the time - I think the same steps should have been taken in this case besides closing outright. None of the people closing the question left a comment. $\endgroup$ – Andy W Mar 24 '16 at 11:50
12
$\begingroup$

I reopened and unlocked the question. The basis for this action is explained in a comment to the question, which points out the (many) statistical issues that need to be addressed. Nevertheless, I think the community acted consistently and reasonably in closing the question, because it is framed in a way that requires detailed knowledge of a particular software application in order to be understood.

When posts can be understood only by knowing software, we should encourage their authors to explain themselves in more universal terms (which, on this site, would be some combination of mathematical notation and English).

$\endgroup$

You must log in to answer this question.

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