-10
$\begingroup$

On August 4, this question was posted. It was perfectly clear to anyone who knows really basic textbook stuff about linear regression but not to those who do not. I added some prefatory basic textbook material and voted to reopen, and then it was reopened.

Now I find this question closed because it is alleged to be unclear. With the one mentioned in the paragraph above, one can see that it was unclear to some because they don't know the basic textbook material, but with this one I cannot so readily surmise why it would be considered unclear.

I don't vote to close questions when I know that the reason why I don't understand them is that I don't know the subject area that the question is within. At this point I suspect that some other users do just that when the subject matter is elementary statistics.

Is there a policy that elementary statistics is in this regard to be treated differently from, for example, algebraic geometry or category theory?

$\endgroup$
11
  • $\begingroup$ Two questions: "What is $\frac d{dx}(x^2)$? And "What is (x^3)' " is also perfectly clear to anyone who knows first semester calculus. But it does not a good question make. $\endgroup$ – amWhy Aug 18 '20 at 19:37
  • 1
    $\begingroup$ @amWhy How does your comment above apply to this question? $\qquad$ $\endgroup$ – Michael Hardy Aug 18 '20 at 19:38
  • $\begingroup$ math.stackexchange.com/questions/3791865/… lacks details, and context. I voted to close as lacking context. Please avoid trying to persecute users by accusing them in manners which are insincere. $\endgroup$ – amWhy Aug 18 '20 at 19:39
  • 2
    $\begingroup$ @amWhy : You want routine details on simplifying the logarithm when that was not what the question is about and when the poster did it correctly? There was a conceptual misunderstanding that was stated clearly. The poster did not understand what was and what was not to be considered "constant" in the context of the problem. $\endgroup$ – Michael Hardy Aug 18 '20 at 19:41
  • $\begingroup$ No, elementary statistics is not more alien than other introductory college or high school math topics. (It may be more boring to some, but that is not the issue.) I'm glad you were able to add a bit of context to the first Question linked. It will make the content more accessible to future Readers. $\endgroup$ – hardmath Aug 18 '20 at 20:26
  • 8
    $\begingroup$ As posters often ask for people to explain their downvotes, I downvoted this post for "At this point I suspect that <list of names>....". If you seriously want to discuss the standards used for closing questions, and look to changing them, you should probably avoid getting personal. It makes people feel attacked, and generally raises the amount of GRAR in the discussion. $\endgroup$ – JonathanZ supports MonicaC Aug 18 '20 at 20:56
  • 3
    $\begingroup$ I can see no real reason for the second question to be closed. The first question, however, is a PSQ and was correctly closed as such (that is, it was not closed as people did not understand the topic, but because it was a low quality question). In a world where take home exams have become much more important, I think it is reasonable to ask users not to answer PSQs. In theory I agree with @hardmath that adding in context is a good thing, but not on the same day that the question was asked. Make the OP work for their exam; after the deadline has passed we can make the question nice. $\endgroup$ – user1729 Aug 18 '20 at 20:56
  • 6
    $\begingroup$ A couple of points before I write an actual answer: It can be helpful to link the exact version of the question that was closed instead of the current one; for you first post for example the post that one sees when following the link is not the one that was closed. While the history is apparent from the revision history this is not clear to everyone and can lead to confusions. It is not clear what the point in mentioning a couple of users is, this is unlikely to be helpful in any way. [Added: I edited the meta post.] $\endgroup$ – quid♦ Aug 18 '20 at 20:56
  • $\begingroup$ @user1729 : It was not closed as a PSQ; it was closed because it was held to be unclear. $\endgroup$ – Michael Hardy Aug 18 '20 at 22:20
  • 6
    $\begingroup$ "Is unclear or lacks details" @MichaelHardy. Please fully explain the close reason given. Lacking details, e.g., context, is sufficient grounds to close. Many users use this reason even for questions lacking context and details. $\endgroup$ – amWhy Aug 18 '20 at 22:46
  • 2
    $\begingroup$ I think the new policy may be helpful to you for how to deal with this situation. $\endgroup$ – Alexander Gruber♦ Aug 19 '20 at 17:36
5
$\begingroup$

To answer the question as asked: There is not policy specifically for elementary statistics questions. It can happen that questions that use some unexplained jargon or make implicit assumptions, can fare better or worse depending on the background of those that first looked at them. There likely is some non-uniformity over the fields. It seems plausible that elementary statistics actually do not fare well in that regard.

That said, I think this is really orthogonal given the examples you gave.

It is a misconception that questions can or should only be closed as "missing context" or "needing clarity" if they are strictly incomprehensible.

It is perfectly possible to understand something but to consider the presentation as unclear or otherwise poor, e.g., making it harder to understand than necessary.

For the first example you give this is rather egregious:

Coefficient of Determination and Standard Error of the Model

i need some help.
$\hat{Y} = 5+2X$
F(Stat.) = 25
$n=102$
$\overset{n}{\underset{i=1}{\sum}}{(Y_{i}-\bar{Y}})^2 = 10$
i got this outputs but i need to find Coefficient of Determination and Standard Error of the Model with using this outputs.

In this form the question has no permanent place on the site. If you or anyone wants to improve it, alright. Otherwise we are better of without it.

For the second example, it is not as bad as the first but first there are only links to images instead of actual formulas and more importantly the poster asks about an error or discrepancy of their calculations with a solution, while not giving their actual work. That's really not ideal.

To reiterate, low quality questions can and should be closed. If they are improved, great, if not the site is better of without them.

$\endgroup$
11
  • $\begingroup$ You wrote "It is a misconception that questions can or should only be closed as 'missing context' or 'needing clarity' if they are strictly incomprehensible." I don't know that any (including me) has ever thought questions should only be closed as "missing context" or "needing" clarity if it is strictly incomprehensible, so your point in bring that up remains unknown. However, how can "strictly incomprehensible" apply to this when it's a question that would be understood by anyone who knows the basics of linear regression? $\endgroup$ – Michael Hardy Aug 18 '20 at 22:25
  • $\begingroup$ . . . and for that matter, how can your comment be relevant when "needing clarity" is just what was cited as grounds for closing it? $\endgroup$ – Michael Hardy Aug 18 '20 at 22:26
  • 1
    $\begingroup$ Is the sloppiness of the style of language in which it was expressed the thing that makes you consider this to be of low quality? $\endgroup$ – Michael Hardy Aug 18 '20 at 22:27
  • 6
    $\begingroup$ @MichaelHardy yes a problem is that the question is presented in a sloppy way that makes it harder to process than necessary. $\endgroup$ – quid♦ Aug 18 '20 at 22:46
  • 2
    $\begingroup$ I do not understand your first remark; turned around I said that it can also be closed as "needing clarity" when it is not incomprehensible (but merely badly presented). $\endgroup$ – quid♦ Aug 18 '20 at 22:48
  • $\begingroup$ But in this case the question is so simple that being hard to process seems not to be something that could happen. $\endgroup$ – Michael Hardy Aug 18 '20 at 22:51
  • 3
    $\begingroup$ @MichaelHardy it uses quite a bit of unexplained notation and terminology. You might claim it is standard, but I disagree with that. I agree that one needs to draw a line somewhere, but I do not think it ever is not to explain anything at all. $\endgroup$ – quid♦ Aug 18 '20 at 23:15
  • 1
    $\begingroup$ What is the nature of your experience with this topic? In those freshman-level courses called "finite math" a bit of probability is covered; will you then claim that something in the theory of probability is not standard because it is not covered in those courses? Certainly this notation and terminology about which you disagree with the claim that it is standard is covered in undergraduate-level introductions to linear regression. $\endgroup$ – Michael Hardy Aug 19 '20 at 1:47
  • $\begingroup$ Look at this question. The one we've been talking about above relies on things covered in standard undergraduate courses, but do you see people reacting to this question by saying it should be closed because the terminology is not explained? $\endgroup$ – Michael Hardy Aug 19 '20 at 2:26
  • 1
    $\begingroup$ If the terminology and notation involved is not standard, then how do you explain the fact that I understood it? Or is it your claim that I did not? (At any rate, it is crystal-clear that you do not know what is and what is not standard in this area). $\qquad$ $\endgroup$ – Michael Hardy Aug 20 '20 at 3:44
  • $\begingroup$ @MichaelHardy I did not phrase that carefully. I disagree that it is not a problem as soon as the notation is standard. I do not disagree much with the assertion that it is standard. I find it especially problematic for notation since it is harder to look up. As for my personal competence on the subject, in principle, I know or at least knew such things; if I had too, I claim I could have figured out what was going on, what I found most confusing is the F(stat.). I also not that you started your answer with "I am assuming..." which suggests that maybe the Q was not crystal clear in the end. $\endgroup$ – quid♦ Aug 20 '20 at 11:52

You must log in to answer this question.

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