12
$\begingroup$

For example: "Why are small sample sizes better?" instead of "Why do small sample sizes give more significant p-values for rare events?":

I tried this just as an experiment because I wasn't getting answers to other, more specific, technically phrased questions. It got more interest and some really good, helpful answers, however I didn't feel great about doing this "trick", and was asked politely (but quite rightly) to rephrase the question which I then did:

Why do small sample sizes give more significant p-values for rare events?

Is this underhanded, to be avoided, or a fair and common tactic to provoke a response, as long as the question is rephrased if and when requested?

$\endgroup$
  • 13
    $\begingroup$ I say no. Clear and interesting questions are what we seek. Jazzing them up is not a good idea. Good writing and good presentation generally and some evidence of research are among the important qualities of a good question. No spin, no shocker headlines, no electioneering please. $\endgroup$ – Nick Cox Mar 30 '16 at 2:10
  • 2
    $\begingroup$ OK, thanks. It feels better to be straight up, even if it's not as effective. $\endgroup$ – Kelvin Mar 30 '16 at 2:54
  • 1
    $\begingroup$ @Kelvin Effective to what end, though? A snappy title invites a snappy reply, and you might be surprised how politely CV users can put someone in their place. $\endgroup$ – Sycorax Apr 1 '16 at 17:36
  • $\begingroup$ If your goal is to attract attention from specialized people able to answer in a very specific way, you may keep your specific but not-controversial formulation and open a bounty. $\endgroup$ – Manu H Apr 8 '16 at 10:11
  • $\begingroup$ stats.stackexchange.com/questions/206592/… serves as an interesting example. Some reservations from myself (with +s from others) on style, but the question has attracted a lot of attention and upvotes and a range of views. Naturally, we don't have a controlled experiment to see how much of that is due to the way it was presented. $\endgroup$ – Nick Cox Apr 12 '16 at 8:29
  • $\begingroup$ @Nick Cox - that question wasn't intended to be phrased controversially, and indeed I don't think it is, but I do agree there is a degree of hyperbole in the statement "from the dawn of time". Even so, it does convey what I meant it to convey, and has certainly attracted a lot of upvotes, while that one question has tripled my official reputation score (not that I pay much/any attention to that). More importantly it has generated a lot of really good quality, thoughtful answers, which is what counts. $\endgroup$ – Kelvin Apr 13 '16 at 7:39
  • $\begingroup$ If that's not controversial, I would like an example that is: "the whole of science has become one big fishing expedition based on false or weak hypotheses" is an example. My own preference is for moderate wording. If there were many more postings of the same kind, I think we'd need to think again about policy. As it is, the occasional question like that can be positive rather than problematic. I'm just asking you to recognise a range of views here on good style, and most crucially the answers below which to me imply that your question would have even better toned down. $\endgroup$ – Nick Cox Apr 13 '16 at 8:01
  • $\begingroup$ ... would have been even better ... $\endgroup$ – Nick Cox Apr 13 '16 at 8:16
  • $\begingroup$ I note that you chose to remove the context from my quote: "It seems that ...", as in my own perception. Perhaps think about that when considering policy? $\endgroup$ – Kelvin Apr 13 '16 at 8:28
  • 1
    $\begingroup$ That's part of my argument, on a meta level. If being controversial and making points in the strongest possible form is fair in a question, then why not in discussion too? Selective quotation is a standard rhetorical move. It's one thing to say "I have a serious question, but I am allowed to pose it controversially" but you are then not so convincing if you object to others acting similarly in debate. (I don't think I was unfair in choosing that example, by the way; nowhere was I claiming or implying that every sentence is of that form or has that tone.) $\endgroup$ – Nick Cox Apr 13 '16 at 10:28
  • $\begingroup$ No worries. In any case I have already agreed with other comments here that controversially phrases questions run the risk of alienating some readers, but I don't see any problem in stating my opinion in what science seems to have become. If you do, then indeed I also agree that you have a right to say so. Personally, I would prefer to focus on asking and answering good questions, but I understand now that some people may care more about style. I guess we all just have to learn to compromise a little. $\endgroup$ – Kelvin Apr 13 '16 at 12:08
  • $\begingroup$ Indeed. The distance between stimulating and exaggerated is short, and to change the image, people don't agree where those places are. $\endgroup$ – Nick Cox Apr 13 '16 at 15:46
  • 1
    $\begingroup$ For what it's worth, I think "since the dawn of time" is a legitimate way to spice up the question - where does the argument for multiplicity corrections stop? But the various assertions & opinions in the body of the question about the state of Science detract from its clarity in my opinion. $\endgroup$ – Scortchi Apr 14 '16 at 14:00
18
$\begingroup$

This part time cabinet maker and cross trainer will shortcut your power analysis with this one weird trick - statisticians hate him!

Being provocative or controversial can garner more attention, but more attention isn't necessarily good. (Be thankful you can only see the ones that weren't deleted.)

While a more controversial title may get more people to respond in some way, it doesn't necessarily get what you need -- a high quality answer -- instead it will tend to garner "snappier" answers from driveby-answerers, and from the regulars it will often attract downvotes, close votes, flags ... or if you're really lucky, just edits.

It also doesn't really help the site serve its purpose; this is intended to be a repository of high quality questions and answers, not clickbait.

$\endgroup$
  • $\begingroup$ All good points, thanks. I do wonder, though, about your last statement: If phrasing questions controversially generates better, more helpful answers (which has been my experience based on a sample size of just 1), then why isn't this serving the purpose to generate "a repository of high quality questions and answers", given that the question can always be rephrased later to match the quality of the answer(s) given? Doesn't the end justify the means? $\endgroup$ – Kelvin Mar 30 '16 at 11:36
  • 2
    $\begingroup$ @Kelvin Glen_b can speak for himself but the idea of a sensational title to be deflated later is disrespectful to the forum, no doubt unintentionally. It implies that we need the inducement of icing on top, but that it can be scraped off when we've performed as wanted. (Incidentally I don't trust these promises to edit later, particularly on behalf of anyone else you can't control.) You don't have the answer you wanted; pushing it again makes the whole idea seem even more inappropriate. Being straightforward with the forum in one's postings is essential as a matter of trust all round. $\endgroup$ – Nick Cox Mar 30 '16 at 11:53
  • $\begingroup$ @NickCox As a man of integrity, I do agree. But as a scientist I also like to challenge assumptions, hence this question. Anyway, the question has been answered, and the answer sits well with me. Thanks. $\endgroup$ – Kelvin Mar 30 '16 at 11:58
  • 4
    $\begingroup$ @Kelvin There's something to be said for an interesting or engaging title; at least, as long as it still serves as a suitable title. The examples in my answer will hopefully help us distinguish engaging from sensational. $\endgroup$ – Glen_b Mar 30 '16 at 12:45
7
$\begingroup$

Glen_b provides a wonderful answer. To date, though, I don't see a key point expressed anywhere in this thread:

A question that makes controversial or incorrect assertions is a confusing question.

Basic logic--the Propositional Calculus--explains why this is problematic. When somebody makes a sequence of assertions $P_1, P_2, \ldots, P_n$ that seemingly imply $Q$, and then asks about the truth value of $Q$, that truth value depends on the truth value of its antecedents (the $P_i$). If any single one of them is demonstrably false, then $Q$ follows logically. A question phrased in such way thereby becomes an uncertain question: does it ask about $Q$ or is it really asking about the $P_i$?

OK, that explanation was too abstract. Let's consider an example. On the future SE health fads site [HealthFad.SE], this question could appear:

Eating disgusting food is good for you. Should I be eating lots of kale haggis?

How should we answer? Should we point out that eating disgusting food is not necessarily a good idea, or should we be addressing the specific merits of kale haggis? Such questions are inherently ambiguous. We should vote to close them until they have been clarified.


If you can successfully phrase a question in some appealing manner, then by all means do so. Attention is good: it gets answers, it generates interest, and it can attract helpful people to the community. But please don't do that in a way that degrades the quality of the question.

$\endgroup$

You must log in to answer this question.

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