Not signed in (Sign In)

Vanilla 1.1.9 is a product of Lussumo. More Information: Documentation, Community Support.

    • CommentAuthorWill Jagy
    • CommentTimeOct 3rd 2011 edited
     
    I put the text below in the interesting thread
    http://tea.mathoverflow.net/discussion/1156/our-standards-and-how-we-enforce-them/
    as a proposal for another section in How To Ask. Then I wondered whether anyone would notice I had made such a proposal, hence the current thread.

    Proposed for How To Ask:


    **Consider asking your peers to help with formulating your question**
    MO is interested in attracting more questions from graduate students and postdocs. We are reliably informed, however, that many are fearful of being treated harshly as MO newcomers.

    One approach, available in existing university programs, is to discuss a potential question with one's colleagues before posting. A group of students might well be able to suggest improvements in the formulation of a question. Meanwhile, a person posting may feel less exposed, knowing that peers have contributed to the post. For someone accustomed to working alone, such personal discussions may be a valuable introduction to the world of give and take that is embodied in seminars and in the journal referee process.

    In another vein, an excellent way to find out what makes a good MO question is to answer several, over the course of a week or two. Answers could be composed with one's colleagues as well. Note that several groups contribute solutions to the problem section of the M.A.A.'s American Mathematical Monthly. One might also answer questions at math.stackexchange.com.
    • CommentAuthorgrp
    • CommentTimeOct 4th 2011
     

    I like the concept, but would suggest more conservative wording, that would not commit MathOverflow to being informed or knowing about university programs. E.g. "We are concerned that fear of looking silly or receiving unsympathetic treatment prevents many such from participating." as a substitute for the second sentence, and "We encourage discussing a question with your peers or your community before posting it on MathOverflow." as a replacement for the third sentence. If a more direct tie with the referee process can be given, then I would include that; otherwise leave it out.

    If a moderator agrees to publish a section of this kind, I can contribute a version that may fit the style of the "How To Ask" page even better than Will's contribution.

    Gerhard "Ask Me About Rewriting Paragraphs" Paseman, 2011.10.04

    • CommentAuthorWill Jagy
    • CommentTimeOct 4th 2011 edited
     
    That's fine, Gerhard. If used, this can also be a group effort, with several drafts, as i am suggesting for MO questions.

    I wanted to give people some reason for coming out of their shells just a little, even when that is uncomfortable for them. I did not think of it that way at the time, but this would have been good advice for me in grad school, then again not something I would have been likely to do as an undergraduate.
  1.  

    Newer users may not be aware of this, but in the old days of MO there was a group of graduates at ... was it Berkeley? ... who had a "twenty questions" seminar where they'd think about some questions that had come up (maybe in their studies). Some of these got posted on MO - search for "twenty questions", there may even have been a related tag. If memory serves me right, Andrew Critch was involved (hope I'm not revealing state secrets here).

    I don't know what happened to it, but I think that such graduate collectives are a great idea in general and having a collective MO presence for one could be a useful addition to such a group.

  2.  

    The twenty questions seminar was started by Andrew Critch and Pablo Solis here at Berkeley. It was a wonderful idea, that launched at about the same time as MathOverflow itself. My impression, however, is that to some extent MathOverflow killed it!

    Still, I think it's a wonderful idea, and perhaps a group of students somewhere wanting to try again could more explicitly think about its relationship with MathOverflow. In particular, at each meeting, after some new questions have been presented, the group could go through them, answering some, being confused or unexcited about others, and perhaps choosing some as appropriate for MathOverflow. This would provide a great setting for polishing and pre-vetting a question, as well as providing some "safety in numbers" if the question is posted under a pseudonymous account created for the seminar.

  3.  

    To borrow a name from a recent book, they could call themselves Nicholas Banacharski ;P

    • CommentAuthorWill Jagy
    • CommentTimeOct 5th 2011 edited
     
    Well, this justifies my long term use of MathOverthrow.
  4.  
    Dear Will,

    I am afraid the title for this thread is most unfortunate, as that combination of words invariably evokes Jonathan Swift.
  5.  

    Alexander: I suspect that was Will's intention, in jest.

  6.  

    Off-topic, but mentions of the 20 questions thing (which I think I first read about on the SB Seminar blog) reminds me that I think someone asked there about the spectrum of "the Scottish flag matrix". Did anyone find a reference which gives a proof that the spectrum does indeed lie on the `saltire' as pictures suggest?

  7.  

    "an excellent way to find out what makes a good MO question is to answer several, over the course of a week or two"

    Seriously, let's not increase people's fears of looking like a noob on MO. Not many grad students are able to answer several MO questions over the course of a week or two on a regular basis. Certainly it takes much more skill than asking a good question. Actually I think that it is the "before I start asking questions, I ought to contribute something to the community / prove myself by answering others' question" attitude that keeps many people from joining MO, including many whose questions would be of much use here. No need to reinforce this attitude, methinks.

    • CommentAuthorWill Jagy
    • CommentTimeOct 6th 2011
     
    Darij, if that is not going to work, I'm not sure how to rescue the paragraph. The bottom line would be some not entirely passive reading of questions by other people, making one's own judgments about the quality of the presentation rather than the nature of the mathematics. We are trained to quickly understand many of the mathematical aspects from a cursory inspection, this is a different task.
    • CommentAuthorWill Jagy
    • CommentTimeOct 6th 2011
     
    Make this a separate item. There may be a way to fit this in: there is a famous American book called "Zen and the Art of Motorcycle Maintenance" by Robert Pirsig. The relevant section is where he is teaching a college rhetoric course (the writing of essays, speeches, etc.). One year, he does something unusual for the first two weeks. Before each class meeting, he chooses four short essays about roughly the same topic. He carefully assesses these with his own substantial training in the field. When class meets, he has the students silently read the essays, and then simply vote, each ordering the essays as to quality, numbers 1-4. The wonderful thing was how close the votes came to his own judgments. He emphasized to the students that they already knew what made a good essay. Eventually they started saying "fine, but how do we do that ourselves?" So he finally began with them to study the textbooks on rhetoric, grammar, writing their own essays.

    Any grad student or postdoc, and many undergraduates, know a poor question when they see it, or at least when an expert has left comments about weaknesses in the formulation. I would like newcomers to lurk for a while in a not entirely passive mode, making judgements about questions they may read, as to good questions and poor questions.

    I can see Darij's point, I am trying to find a milder version of this general idea...
  8.  

    @Will: I think that the point you are trying to convey (at least in the way I read it) is true for pretty much every on-going community that a person wants to join. One should observe, for a day or two at least, the community and review its behaviour and so on.

    Only after being able to somewhat assess the norms of this community one should join in.

    • CommentAuthorWill Jagy
    • CommentTimeOct 6th 2011
     
    It would appear that the second part of my draft is converging to the null set of advice. So, we are back to possibly including, in How to Ask, some less-insulting variant of

    **Consider asking your peers to help with formulating your question**
    MO is interested in attracting more questions from graduate students and postdocs. We are reliably informed, however, that many are fearful of being treated harshly as MO newcomers.

    One approach, available in existing university programs, is to discuss a potential question with one's colleagues before posting. A group of students might well be able to suggest improvements in the formulation of a question. Meanwhile, a person posting may feel less exposed, knowing that peers have contributed to the post. For someone accustomed to working alone, such personal discussions may be a valuable introduction to the world of give and take that is embodied in seminars.


    In particular, Scott M. has some history in mind with: "Still, I think it's a wonderful idea, and perhaps a group of students somewhere wanting to try again could more explicitly think about its relationship with MathOverflow. In particular, at each meeting, after some new questions have been presented, the group could go through them, answering some, being confused or unexcited about others, and perhaps choosing some as appropriate for MathOverflow. This would provide a great setting for polishing and pre-vetting a question, as well as providing some "safety in numbers" if the question is posted under a pseudonymous account created for the seminar."
    • CommentAuthorquid
    • CommentTimeOct 7th 2011
     

    I appreciate this effort. However, in details I have some concerns. On the one hand, I share Darij's concern. On the other hand, while I think to discuss things in groups can be a great thing, I am a bit sceptical about making this suggestion explicit in the present form. In particular, I am not so sure that a couple of people that do not know MO, will collectively write a better MO question than a single person. Regarding mathematics, yes perhaps, and some things on the simple end might be catched as somebody will solve it off-line. But regarding style, I doubt it. Moreover, I think that there are some points over which even experienced people can stumble when asking an MO question without being familiar with the medium. While the Zeilberger-incident was too special a case to serve as a representative example, it still somewhat illustrates this. But there were other cases, where experienced people suffered a somewhat unfortunate welcome due to various and varying circumstances.

    A suggestion that I do like is to 'encourage' pseudonmys, for those feeling unsecure, at least for the start (okay, I am biased here). But, to me it really seems, to a large extent, to resolve worries about an unfriendly welcome. On MO one does have a second chance to make a good first impression.

  9.  

    I don't think Will's suggestion is meant to cover every possible anecdotal example, quid. I suspect there is a group it's targeted towards, but beyond that it appears to be illustrative of the expectations the community has of what goes into a good question. This isn't your average forum from the perspective of casualness, and it's only fair to explain that somehow.

    • CommentAuthorquid
    • CommentTimeOct 7th 2011 edited
     

    Ryan, okay, you no doubt understand (North American) graduate students better than I. But somehow I cannot help but feel that this will be a case of what Paul Siegel descibes at the start of his point 2.

    • CommentAuthorWill Jagy
    • CommentTimeOct 7th 2011 edited
     
    Quid, I am getting resigned to the idea that it is unlikely that anything from this thread will make it into How to Ask. That being said, all I really had in mind was finding ways to spread out the anxiety Paul Siegel mentions, while possibly improving some questions. The target population is very specific, graduate students and postdocs, in university mathematics programs where everyone sees each other and face to face discussions are frequent, department tea and the like. Also, I did not envision Scott's regular meetings about all things MO, just one student with one question. In such a case, I admit asking one's friends (or peers, and perhaps not entirely friendly) for assistance is itself a source of potential anxiety.

    This was a response to Paul, in that I probably understand his concerns, but I am not sure how much nicer I am willing to be. Indeed, in the one case (Madalina) where I went overboard trying to rescue a poorly formed question, it turned into a nightmare for all concerned. Better if I had let it go, and let others say it was a poor question. Perhaps the same outcome, of course.

    About the target audience, we do want more questions from grad students and postdocs, and these suggestions are at least possible for those people.

    Anyway, after Ryan's suggestion on Pauls' thread, and Andrew Stacey's argument that grad students are at the perfect level for the questions we want, my interpretation was that I had a responsibility to compose a genuine proposed section for How to Ask, so I did that.
    • CommentAuthorquid
    • CommentTimeOct 7th 2011
     

    Will, I understood the good intent, still I am worried it could have the opposite effect (somewhat along the lines of what Darij said). Also, thank you for the explanation in the other thread.

    My personal impression is that there are mainly a handful of basic errors that people make with their questions that they all (or most of them) could avoid if they were aware of the pitfall. [Perhaps this is not true, and the basic errors are just those things I find unfortunate, but then I have seen a lot of question and others people reactions, so it seems to me there is a pattern.] And, perhaps it could help to record them in a very brief form; only linking to more detailed expositions. In an informal way, some does and dont's for new inexperienced users as I see them (essentially everything is already written somewhere). For the moment three:

    1. Ask a focued mathematical question. And, only if you really care about an answer.

    2. State why you asked the question, why you care about an answer to the question, and give some brief information on your background.

    3. Show that you made some effort when compsing the question. In particular, write/type the answer carefully. There is no need to be overly formal, but write full sentences using proper spelling and grammar (to the best of your ability).

    IMO if everybody would just try to follow this, there would be much less problems.

    • CommentAuthorWill Jagy
    • CommentTimeOct 7th 2011
     
    Quid, those seem pretty good guidelines.
  10.  

    (Sorry for being so late to the party. Things have been a bit crazy lately.)

    I like the idea of adding such a section to the How To Ask page. The page is supposed to be a combination of common sense things that everybody should be doing already (but sometimes forgets) plus bits of wisdom we've picked up about how to use the SE engine effectively. Asking a small audience before asking a bigger one is a good idea for anybody. But rather than focusing so much on the fact that people may be nervous about posting, why not say something to the effect of

    Ask your colleagues. Kicking a question around with a few people will often help narrow down exactly where the problem is. It also helps ensure that the basic premises and the formulation of the question are sound.

    I think something like this may have the desired effect (i.e. give people a pretty good idea of when they can be sure their question will fly) without implicitly reinforcing the idea that you should be terrified to ask any question.

    • CommentAuthorWill Jagy
    • CommentTimeOct 10th 2011
     
    Works for me.