9
$\begingroup$

In the last couple of weeks we have been polling for opinion about a possible list of words (or phrases) that will filter questions out of the HNQ list.

(See this which evolved into a poll and a discussion.)

Some members of the community were extremely critical of the idea, and others were simply very critical of some suggestions. For example, at the time of writing, the word "advice" has a +8/-12 score, "interesting" has +7/-13, and "question" has +5/-13.

I am personally opposed to the use of vague pronouns ("this" has a +6/-15 score at the moment), as I find them to be very uninformative and lacking.

Let me put aside the discussion about the HNQ filter, which is not a ban on titles from the site anyway. Instead let me talk about what I feel is an odd reaction of a significant part of the community to the issue of titles.


I raised an issue, nearly 9 years ago, about subjective titles. How should we treat subjective titles? where I asked what should we do about words like "hard" or "difficult" that appear in titles. Often found in the context of a completely bland title "Difficult integral question" or "Hard inequality".

The issue is not the subjective part. It is the uninformative part. And titles should contain useful information about the question, which mostly subjective words do not add. But also words like "help" or "question" are not adding, and are usually used as a replacement for something that is in fact informative.

Even the arguments in favour of allowing subjective words are not really in favour of allowing them in the title. Yes, there is merit for someone using the term "diffiuclt integral" in the body of the question, where more context can be given. But there is no reason to put this into the title, which should give an idea of what lies inside the question itself.

So I wanted to bring this up, again.

What is a good title?

Should it have some information about the actual question? Should we even care what is the title of a question?

| |
$\endgroup$
  • 4
    $\begingroup$ "I am personally opposed to the use of vague pronouns ("this" has a +6/-15 score at the moment), as I find them to be very uninformative and lacking." I know. The entire world know about it. Think about $+6$ vs $-15$ for a moment, what could one imply from it? $\endgroup$ – allesia_b Apr 14 at 14:49
  • 6
    $\begingroup$ @Magic_Isa: Yes, which is why I am asking the community what do people think is the role of a title, and what counts as a good title. Let's focus on that, yes? $\endgroup$ – Asaf Karagila Apr 14 at 14:57
  • 4
    $\begingroup$ Ok. "And titles should contain useful information about the question, which mostly subjective words do not add." you're right. I think we all agree in this. This would be ideal, but it's not the reality. Keep in mind that some users, specially new users just ask like if they were in a classroom, among friends, etc. They don't know our formalities, but that of course doesn't mean that their question won't be as valuable & important as the one with such a -perfect title- $\endgroup$ – allesia_b Apr 14 at 15:07
  • 6
    $\begingroup$ @Magic_Isa: Yes, new users need to be educated. And experienced users need to educate them. But also one way of doing that is making an effort that when people see questions, they see them with good titles. Can you please stop making this about the HNQ, though? $\endgroup$ – Asaf Karagila Apr 14 at 15:13
  • 3
    $\begingroup$ Asaf, a good title can contain the word "this", so you should not be opposed to the use of it. $\endgroup$ – allesia_b Apr 14 at 15:59
  • 9
    $\begingroup$ @Magic_Isa: Honestly, if you don't have anything relevant to add, it's maybe a good idea to not add anything. $\endgroup$ – Asaf Karagila Apr 14 at 16:27
  • 8
    $\begingroup$ No, no, if you want to delete the comments, you're free to do that yourself. I want to understand how telling me that "this" could appear in a good title (something that I never denied, by the way), is relevant to the discussion on what is a good title. $\endgroup$ – Asaf Karagila Apr 14 at 16:34
  • 7
    $\begingroup$ My point was, that it seems that you don't want this on titles, but it could be good to have it on titles. // Why am I mentioning this? Because you bring the topic in your post. And there's no need to be upset. $\endgroup$ – allesia_b Apr 14 at 17:06
  • 6
    $\begingroup$ The current title of this question ("Question to the community on a problem ") is so general that it could be argued to be clickbait. It's not clear if such a title was purposely chosen in order to make a point about bad titles, but I think it would be be better to instead lead by example and give it a title that you consider to be a good title. Also it would help to mention the meaning of the polarity of the cited votes. $\endgroup$ – Bill Dubuque Apr 14 at 18:36
  • 6
    $\begingroup$ @Gone: When I try to lead by example, I get called out for being a censor, dictator, or otherwise. This time, I follow the example of the community. If the quality of the title is irrelevant to the quality of the question, I decided to raise an issue and I put effort into the question. Do you really care about this title? $\endgroup$ – Asaf Karagila Apr 14 at 18:39
  • 5
    $\begingroup$ @Magic_Isa the argument you gave was anyway backwards. It's not so much that no one will ever have a look, but that it would be better to give more information for the convenience of readers and future organisation. I mean there is no need to have a title on the cover of a book either. Just open it and you'll see. There is no need to sort the books in a library, just search a bit and you'll find. $\endgroup$ – quid Apr 14 at 18:53
  • 10
    $\begingroup$ @Magic_Isa do not flood threads. At some point it becomes too much and it backfires. $\endgroup$ – quid Apr 14 at 18:57
  • 10
    $\begingroup$ @Magic_Isa: You're not the community. $\endgroup$ – Asaf Karagila Apr 14 at 19:49
  • 6
    $\begingroup$ @Theo: I think that people simply forget the titles. Often. You see that a lot in edits, someone edits the post to perfection, but leaves the title out. I think that we can make this a conscious thing, and it will improve, at the very least with those people aware of the problem. $\endgroup$ – Asaf Karagila Apr 15 at 18:07
  • 4
    $\begingroup$ @AsafKaragila I am certainly guilty of trying to fix up a post, but neglecting the title. I've even done this when my goal in editing was specifically to clean up a title! $\endgroup$ – Xander Henderson Apr 15 at 18:59
21
$\begingroup$

What makes a good title?

Good titles should unambiguously tell the user what they are getting into before they click on it. The goal is to attract the attention of those people who have the expertise to answer the question or may be interested in an answer, while not wasting the time of readers who don't care. Perversely enough, a good title might decrease the number of views, but should increase the average quality of interaction (as measured by time spent reading the question, ability of the reader to answer, and/or other folk having a similar question).

Good titles will generally consist mostly of nouns, verbs, and maybe a preposition or two (where one might regard an expression like $x = y$ as a SVO sentence or sentence fragment). Adjectives and adverbs should be few and far between, as they typically don't add much in the way of content (there are exceptions, particularly when such words are used in a technical capacity, e.g. a question about a "hard integral" is very different from a question about "hard analysis").

Good titles should quickly get to the point of the question. Ideally, a brief version of the question should be stated in the title. Because this website is about asking mathematics question, good titles will generally (though not always) include notation. Notation should be included, and should be typeset correctly[1], whenever it is appropriate.

Good titles include enough detail about the context surrounding the problem to provide signposts for potential readers. For example, the expression "$x^2 + y^2 = (x+y)^2$" will mean one thing if the title makes it clear that the problem was encountered in grade-school algebra, and another if accompanied by a phrase such as "freshman's dream" or "field of characteristic two". Thus a good title for question about this identity should include something about the material being studied. Other context which might fit into a title are a indications of the asker's level of knowledge, the general field they are studying (particularly if tags are insufficient), a text author (there are a large number of questions on MSE with titles containing some variation of "Theorem X from Baby Rudin"), or the provenance of a competition problem.

Examples

To build an example from something I encountered today, let us suppose that I have the following question (which is likely an appropriate question for MESE, but that isn't really the point):

Today, while helping my daughter with her remote-instruction schoolwork, I discovered that they students in her second grade class were being taught to solve linear equations in two variables using a method involving colored bars. The teacher's notes refer to this as the "bar model", but there is no explanation for how it works.

My daughter seemed fairly proficient with the technique, hence it went by too quickly for me to parse what was going on. It seems that she was using various bars of differing length (or color, maybe?---she seemed to be insistent on using different colors of whiteboard markers in particular way) to represent known and unknown quantities in a system of equations. The way in which bars were "stacked" or "concatenated" seemed important, though, again, my daughter worked through the problems faster than I could discern the pattern.

Does anyone know what this method is called, and can anyone explain how it works or point me in the direction of useful references?

So, what title should I give this question? A few possibilities include:

  • What is this crazy thing my daughter is doing?!

    This title is less than useless. It gives the reader no idea what the question is about, and is, at best click-bait. Moreover, it editorializes (via the adjective "crazy"). My guess is that nearly every click on this title will be quickly followed by the closure of a tab or a user hitting the back button.

  • Can someone explain this technique for solving systems?

    This title is better, but still leaves off a lot of context. What is "this" technique? Who is solving these systems? What is the level of student? The basic question is in the title, but there is no context to orient the reader (aside from a vague reference to "solving systems").

  • What is the "bar model" used by an 8 year old to solve systems?

    On short notice, this is about the best I can do. The phrase "bar model" is important context (it was on the teacher's slide, after all), and the inclusion of my daughter's age tells the reader that a tenured professor with research in systems of nonlinear PDE probably doesn't care to click through, while an expert in grade-school education immediately knows that they might have an answer.


[1] For technical reasons, there is a small downside to including MathJax in titles: questions with MathJax'd titles cannot appear in the "Hot Network Questions" list, because the MathJax renderer is not loaded for every site in the network. To avoid ugly code in the titles of questions rendered on other sites, such questions are excluded from the list. However, this technical concern about how mathematics questions are advertised to the wider non-mathematics community should be a minor, secondary concern (in my opinion).

| |
$\endgroup$
  • $\begingroup$ Thank you Xander, for a solid answer that addresses the issues raised. $\endgroup$ – JonathanZ supports MonicaC Apr 15 at 5:05
  • 2
    $\begingroup$ Thank you! Now I know that every good question should have some MathJax in the title, as this has the upside of keeping it off the HNQ list, thereby avoiding the danger of having a horde of under-informed people come come along and up-vote the bad answers, down-vote the good ones, and add noise to the comments. Only a small upside, because good questions are not in much danger of making the HNQ list anyway. $\endgroup$ – bof Apr 15 at 5:26
  • 4
    $\begingroup$ +! for the first sentence! Titling a question is like titling a novel -- and just as hard. $\endgroup$ – postmortes Apr 15 at 5:27
  • 6
    $\begingroup$ (+1) This seems like solid guidelines/advice for titling questions. A minor comment though: Imho, the use of imperative in asking questions comes off as rude, so in the best title example, I would probably prefer either "Could someone explain the 'bar model' used by an 8 year old to solve systems?" or just "A 'bar model' used by an 8 year old to solve systems". $\endgroup$ – jgon Apr 15 at 7:02
  • $\begingroup$ Nice answer. Very nicely written. $\endgroup$ – user170039 Apr 15 at 8:36
  • 6
    $\begingroup$ @bof I am not sure if I should be reading your comment as sarcastic or sincere, but in either case, I think that the nature of "good title" is (and should be) orthogonal to the HNQ. I understand that this discussion was prompted by possible changes to the HNQ, and I get that there are strong opinions and emotions on both sides of that issue, but I really want to try to address Asaf's question in isolation from those issues. Another related question which I also don't think is pertinent to this conversation: can a question be considered "high quality" if it has a poor title? $\endgroup$ – Xander Henderson Apr 15 at 20:16

You must log in to answer this question.

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