68
$\begingroup$

Sometimes this "What have you tried" thing seems silly.

Take this question [edit: question deleted/link broken] for example. It's obvious that one should try to do it by induction, but it's not the same as asking how to prove $\sum_1^n j=n(n+1)/2$, in that it's not obvious, or at least it wasn't obvious to me, how to make the proof by induction work.

Supposing that's so, what would "I tried to do it by induction but I don't see how" have added? I gather that that would make the question more acceptable, but I really don't get it.

Edit: I've been told that's a senseless question because the answer is obviously "that would have added nothing". It was a rhetorical question, sorry.

Next question: Fine. Then what specifically could the OP have said, about that question, regarding what he'd tried, in order to show he'd worked on it?

That's rhetorical as well, although much less so, certainly less obviously so. I can't imagine anything specific the OP could have said along those lines. Here's why: I did spend a few minutes "working" on the question. Wasn't going to give it a lot of time, but I did work on it, cuz it pissed me off, seemed like something I should be able to do. But during the time I was working on it it didn't happen once that I tried something that failed. It just wasn't that sort of difficulty - where I was stuck was thinking of something to try.

It's a fact that I did work on it, and it's a fact that if I hadn't seen the solution I'd be totally unable to give any evidence whatever that I'd worked on it.

My point, for those to whom it's not clear, carefully phrased as a question: Might it not be a good thing if before whining about "show your work" we stopped to consider that there is such a thing as working on a question without being able to produce evidence to that effect?

Supposing one agreed with that in principle, one could try to avoid asking unreasonable "what have you tried?" questions by making certain that before asking that one actually had at least a reasonable approach in mind. (In the question I'm talking about the guy who said "Try induction" more or less admitted when asked that he didn't see how to do it by induction. Suggesting induction as something that might work, fine, bbut suggesting as though it was clear it was something the OP should have tried but didn't, not fine unless we do see how it works.)


(So as to avoid answering questions from someone who really has no clue and hasn't even considered induction, I posted an answer which it seems to me should help only if the OP is thinking about induction...)

$\endgroup$
  • 30
    $\begingroup$ So what have you tried to do to solve this problem? Have you tried induction? :) $\endgroup$ – Asaf Karagila Mar 18 '18 at 18:18
  • 14
    $\begingroup$ Not everyone who does mathematics is a man. Consider using gender neutral language such as "they" and "person" instead of "he" and "sir" when addressing or talking about other users. $\endgroup$ – Namaste Mar 18 '18 at 18:34
  • 46
    $\begingroup$ @amWhy You can locate the question by noting which sentence ends with the character "?". $\endgroup$ – David C. Ullrich Mar 18 '18 at 18:45
  • 10
    $\begingroup$ Sarcasm isn't going to get you far. And if that's your only question, that's senseless, and a misrepresentation of the question. $\endgroup$ – Namaste Mar 18 '18 at 18:57
  • 34
    $\begingroup$ @amWhy Sarcasm? I had a question. I indicated it was a question using standard punctuation, and you nonetheless asked what the question was. Senseless? I don't see how a _question can be senseless. Nobody's said anything remotely like an answer. The question was the question - I don't see how it cann be a misrepresentation of itself. $\endgroup$ – David C. Ullrich Mar 18 '18 at 19:13
  • 15
    $\begingroup$ A few years ago I suggested this. The gist is that anyone yelling WHYT should themself be able to answer the question fully, without breaking a sweat. That proposal never won much support. I guess that problems with the proposal are A) timing (speed is essential when closing a question because we have those answering machines), B) verification (may be having a sufficient number of upvotes in a relevant tag would be a good enough lithmus test?) $\endgroup$ – Jyrki Lahtonen Mar 18 '18 at 19:42
  • 24
    $\begingroup$ @amWhy You address borderline insulting comments to me, and when I simply reply I'm "targetting" you? Wow. I mean really, wow... I take it that no, you can't give a specific example of something the OP could have said about that specific question in order to "show some effort"? $\endgroup$ – David C. Ullrich Mar 18 '18 at 22:05
  • 18
    $\begingroup$ My own concern with "what have you tried" is that it is only part of what I really want to know, and often the least important part. I want to know why a question is interesting, if it isn't in a field I already know. I want to know where it came from and what problems it is related to. And, finally, I want to know how the OP thinks about the question and what methods they have at hand. I find I'm not always happy with questions that include a token "My effort" without adding genuine context. I try to avoid asking "what have you tried", instead I like to link to "how to ask a good question" $\endgroup$ – Carl Mummert Mar 18 '18 at 22:16
  • 8
    $\begingroup$ In February, several editors also added advice on "Avoid no-clue questions" math.meta.stackexchange.com/questions/9959/… to the canonical "how to ask a good question" post, $\endgroup$ – Carl Mummert Mar 18 '18 at 22:18
  • 11
    $\begingroup$ This is redundant for people who were around for some time, but maybe it is worth reminding that there are various ways to add context, it does not necessarily mean OP's own attempt: How to ask a good question. and Can we stop the “Show your work craze”? $\endgroup$ – Martin Sleziak Mar 19 '18 at 1:26
  • 21
    $\begingroup$ @amWhy As I earlier replied to a similar comment by someone else, no one is actively suggesting that everyone who does math is a male. It's just that we're accustomed to such usage. I'm an atheist, and I say 'Oh my God!' all the time. $\endgroup$ – MalayTheDynamo Mar 19 '18 at 10:38
  • 9
    $\begingroup$ Once again, a closure vote is not a super-disagree button. I'm voting to reopen because it is perfectly obvious what is being asked here, regardless of whether you like or dislike the premise. $\endgroup$ – T. Bongers Mar 19 '18 at 16:02
  • 7
    $\begingroup$ @user296602 Right - for some reason I didn't see it was on hold until I refreshed the page. Yes, the idea that it's unclear what I'm asking is silly - good luck with that. It does seem curious that such a bad bad question should have so many more upvotes than downvotes. Raises the question of whether the upvoters are evil or just ignorant. $\endgroup$ – David C. Ullrich Mar 19 '18 at 16:36
  • 7
    $\begingroup$ @user296602: I did not vote either way, but I do not think it's "perfectly obvious what is being asked here". For example, most (all?) of the interrogative statements in the posting are rhetorical, not questions meant to be answered. $\endgroup$ – Hurkyl Mar 19 '18 at 18:51
  • 11
    $\begingroup$ This comment thread is as entertaining as it is predictable. I love seeing WHYT questions in the Hot Meta Posts area. $\endgroup$ – tilper Mar 20 '18 at 22:40
52
$\begingroup$

I have taken, recently, to broadening my meager repertoire of responses to context-free questions beyond a bare "What have you tried?" because, in essence, I agree: There are plenty of times when that question just doesn't seem sensible. The hardest task is occasionally just figuring out how to begin. Alternatives to WHYT that sound reasonable to me include:

  • What are you studying?
  • What text is this drawn from, if any? If not, how did the question arise?
  • What kind of approaches (to similar problems) are you familiar with?
  • What kind of answer are you looking for? Basic approach, hint, explanation, something else?
  • Is this question something you think should be able to answer? Why or why not?

This list is not meant to be exhaustive, nor are all of them applicable to all questions, but I have found them useful options to keep in mind.

$\endgroup$
  • 5
    $\begingroup$ Right, the goal is not to humiliate the user who is generally confused about how to proceed, nor to drive them from the field, but to develop a rapport sufficient to achieve mathesis (now rare). $\endgroup$ – hardmath Mar 20 '18 at 22:14
  • 1
    $\begingroup$ These are good alternatives. For some reason I can't quite articulate, "what have you tried?" seems to me appropriate only for straightforward questions or something similar to that (this impression seems to be shared by some of the community, from what I could gather from the comments). Your suggestions are more broadly appropriate (some more than others, of course). $\endgroup$ – Aloizio Macedo Mar 20 '18 at 22:34
  • 21
    $\begingroup$ Furthermore, they circumvent the issue that David points out: which is that "there is such a thing as working on a question without being able to produce evidence to that effect" (which I agree with). "What have you tried" seems to ask for effort, which sometimes is just noise and is arguably mostly an ethical requirement, whereas your suggestions seem to ask for context, which is what most of us can agree is the most important, I believe. $\endgroup$ – Aloizio Macedo Mar 20 '18 at 22:38
  • $\begingroup$ And consider also just helping them. Spending too much time trying to get context you disregard the value of the question at face value. $\endgroup$ – AmateurMathPirate Mar 21 '18 at 13:19
  • 3
    $\begingroup$ Actually addressing the point I was trying to make, instead of replying as though I'd been insisting that every question should be given a complete and detailed answer regardless of "quality of question" issues - wow. $\endgroup$ – David C. Ullrich Mar 21 '18 at 13:46
  • 2
    $\begingroup$ @AloizioMacedo ' For some reason I can't quite articulate, "what have you tried?" seems to me appropriate only for straightforward questions': Whether or not you can articulate the reason, that's an excellent articulation of what I was trying to say. In extreme cases the reason seems clear - if I have no idea what one might try then asking the OP what he or she has tried seems obviously inappropriate. $\endgroup$ – David C. Ullrich Mar 21 '18 at 13:49
  • 2
    $\begingroup$ @hardmath: Have you not yet realized that the rarity of mathematical thinking nowadays is precisely because mathematics students can easily cheat by outsourcing their homework and getting full solutions? $\endgroup$ – user21820 Mar 21 '18 at 14:39
  • 14
    $\begingroup$ @user21820: That's plausible, and certainly tempting to believe, but for that very reason I distrust it; it has a bit too much of the "kids these days" feel. It's almost impossible to demonstrate in any really convincing way, I think. Are we even sure that it's rarer, or is it just that we're confronted more day after day by people who lack it? ¶ In any case, we can only deal with the situation in front of us. I think WHYT is designed to elicit that sort of mathematical thinking; the challenge with it is often that it simply doesn't provide enough of a nudge in the right direction. $\endgroup$ – Brian Tung Mar 21 '18 at 15:46
  • 2
    $\begingroup$ @BrianTung: I personally don't use "what have you tried". However, I have completely no sympathy for serious cases of super-'entitled' askers like this. As for convincing, I'm convinced because I've been on Math SE for a long time. I don't expect to convince others, but anyone is free to dig up the data and perform a statistical analysis. $\endgroup$ – user21820 Mar 21 '18 at 16:09
  • 1
    $\begingroup$ Oh and when I was doing my undergraduate I personally witnessed many classmates outsource their homework to Yahoo Answers. This simply wasn't possible in the days before the internet. $\endgroup$ – user21820 Mar 21 '18 at 16:11
  • 2
    $\begingroup$ @user21820: Of course. I don't have much sympathy (if any) for such questions. I'm convinced that such people exist, but I'm also convinced that they've always existed. What I'm not convinced of (in either direction, yea or nay) is whether what we see is indicative of a general decline in mathematical thinking. Most students still do not know of Math SE. It's still pretty self-selecting. $\endgroup$ – Brian Tung Mar 21 '18 at 21:21
  • $\begingroup$ In line with this answer, I would suggest to use something along the lines of "How did this question arise? Were you able to make any progress on it?" (Note how it is totally natural to reply to the second question that one did not manage to make any progress despite $x$ time units spent.) $\endgroup$ – Lord_Farin Mar 21 '18 at 21:44
  • $\begingroup$ +1 for the points you have raised. For the few questions I have asked here I have actually not shown my work, but rather added context as given in your points 2,3,4. $\endgroup$ – Paramanand Singh Mar 22 '18 at 1:22
  • $\begingroup$ A possible context guess or short hint could help the OP more than a dry WHYT greeting. $\endgroup$ – Narasimham Mar 22 '18 at 14:49
  • 1
    $\begingroup$ I sometimes suggest the OP work out a simpler analogous problem.or a special case. This is often appropriate for combinatorial questions. $\endgroup$ – Ethan Bolker Mar 31 '18 at 0:17
59
$\begingroup$

The underlying problem, in my opinion, is the conception that one of the primary goals of MSE is

Student arrives with an exercise and leaves with a solution.

In other words, one posts to MSE in order to outsource their mathematical exercises.

But this is (again in my opinion) an inappropriate use of the site! What we really want is something more like

Student asks a question about mathematics and learns something, and maybe others learn too!

It's fine if an exercise prompted the question, and it's great if the student's new knowledge allows them to solve their exercise and future exercises as well, but the important thing is that the point of the interaction is to learn something mathematical!


The point of WHYT (in my opinion) is that if questioners post enough about their knowledge and thought processes, that answerers can divine the questions that should have been asked — the questions about learning mathematics that the questioners need — and proceed to answer that.

In other words, WHYT is meant as an aide. Rather than close a question outright, it is an attempt to help the questioner transform a question that is inappropriate for MSE into one that is appropriate for MSE, or at least implies an appropriate question.

If the only thing a user will provide is an exercise, then the question is simply inappropriate for MSE.

$\endgroup$
  • 14
    $\begingroup$ Incidentally, I share your dislike of WHYT, but for the opposite reason -- I don't think it's effective in its goal, and has the negative side effects of causing grief and perpetuating the conception that MSE is about arriving with exercises and leaving with solutions. But I'm out of touch with the current trends, so I can't claim confidently that its ineffective. $\endgroup$ – Hurkyl Mar 19 '18 at 3:58
  • 4
    $\begingroup$ While I agree that it shouldn't be one, MSE has essentially become a homework help site. Not only do homework exercises not get closed, I usually can't even finish typing a hint for a 5-minute-old question before some whiz kid posts a solution with full details. $\endgroup$ – user360874 Mar 22 '18 at 3:22
  • $\begingroup$ @user37208 Some elders too post solutions very quickly. Example - That Inequality Guy. (No offence) $\endgroup$ – Jaideep Khare Mar 22 '18 at 14:52
  • 2
    $\begingroup$ If there is a natural tendency for MathSE to turn into a homework help site, then insisting that it must not do that is a good way to stay frustrated. The question is how to generate the most value, given that people are going to do what people naturally do. $\endgroup$ – G Tony Jacobs Mar 24 '18 at 0:11
  • 3
    $\begingroup$ @GTonyJacobs: There is a natural tendancy for MathSE to turn into a spam repository too, but we do a pretty good job of fending that off. And it's not like its an impossible task to prevent a stackexchange site from being a do-my-work-for-me site, even one with our volume. Stackoverflow managed it, for example! $\endgroup$ – Hurkyl Mar 24 '18 at 12:42
  • $\begingroup$ @Hurkyl Spot on! See my answer on how to provide a method to help achieve understanding instead of 'answer keys'. $\endgroup$ – Discrete lizard Mar 25 '18 at 13:29
  • $\begingroup$ @Hurkyl (2 comments up) I actually think Stack Overflow has the same problem. It's fairly common that I see someone post little more than an abstract problem description and get an answer with a full implementation; or someone posts a chunk of code that doesn't work, and answers consist of more chunks of code with bugs fixed and the style improved. In both cases I have to imagine that the asker and most readers learn little or nothing from the exchange. $\endgroup$ – David Z Mar 27 '18 at 23:41
  • $\begingroup$ You have nailed it with "WHYT is meant as an aide". As a Stack Overflow denizen who has landed here by accident, I'm happy to see this stance being advanced elsewhere on the network. (By the way, @DavidZ 's observation is also spot on.) $\endgroup$ – duplode Mar 29 '18 at 3:14
19
$\begingroup$

In addition to the fact that we're not meant to do a person's homework for them, there's a more subtle reason to ask a question like that. Particularly for questions where the poster has just copied the homework question, it may not always be obvious what the context for the question is.

Typically, if you've got a homework question, you're supposed to apply something you've recently learned to solve it, but we don't know what class the poster is taking, let alone what material was most recently covered. The poster may be looking for something involving basic arithmetic but their question may also have solutions in Galois Theory, or Combinatorics, or Advanced Hypermathical Topomology, and it's not going to be helpful for the poster to offer the wrong one.

$\endgroup$
11
$\begingroup$

I'm a frequent answer-er and basic homework hint-er on Computer Science. Perhaps some outsider perspective is useful here.

Do note that we have chosen for a very different attitude wrt 'homework' than math.se, for a good reason, in my opinion.

Our way to help the 'clueless' is the following procedure:

  1. If it looks like a verbatim copy of homework, politely tell them they shouldn't just dump homework and expect us to solve it. Us solving their homework is unlikely to help them.
  2. If it seems they have done something but are stuck and merely state confusion, ask them to clarify.
  3. Okay, if it now is clear what the problem is an whether it is trivial, decide whether an answer or hint will help. Err on the side of hinting, answers will spoil the hints, but not the other way around. On CS, we usually hint in comments, as they aren't really answers.
  4. Finally, if it seems this now look like a serious question, answer it!

Ideally, questions start at 1 or 2 and step by step become 3 or 4 and remain there. This takes a bit of back and forth, but seems to work well. Some are really helped by 3 (i.e. the hint has helped them to solve the question by themselves!) or other times they had a real serious question that has merely been badly asked and lead to 4. All this can be done before closing the question on CS.SE, but perhaps this is unsustainable on a site of your size.

Of course, people often get stuck at 1 or 2, but it isn't hard to advance to 3 or 4. It merely means you have to take your question and SE seriously. If you cannot do that then, well, you are better of without us.

$\endgroup$
  • 1
    $\begingroup$ It seems that in math.se a complete solution would be posted when the question was somewhere at stage 1 or 2. Then the question is unlikely to advance to stage 3,4. What would you guys do in CS.se to those questions? $\endgroup$ – user99914 Mar 25 '18 at 14:22
  • 5
    $\begingroup$ @JohnMa This situation is indeed common on CS. At times, we give up and do nothing, but sometimes we post Please consider not to encourage [undesirable posting behaviour](https://cs.meta.stackexchange.com/questions/594). $\endgroup$ – Discrete lizard Mar 25 '18 at 14:41

You must log in to answer this question.

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