Is it ever appropriate to share one's thoughts on a topic as a way to garner constructive feedback and opinions from the community instead of asking a direct question, or is there another place for this? Example:

The following is my understanding of the merits of logistic regression vs SVM for classification.

  1. Point 1
  2. Point 2
  3. Point 3

Can anyone correct or add to my understanding?

Now I know that this is technically a question, but the spirit of it is somehow different than traditional forum posts which are seeking to solve a specific thing.

  • 3
    I would rather comment on an example than discuss the principle of where to draw a boundary. Generalizing from your specifics, it can indeed be extremely difficult to answer when the post is of the form "I don't understand X", which is, strictly, a statement, not a question. But here's a positive suggestion. You have a question, which you can immediately follow with your own (tentative) answers. Then other people will be able to edit, comment and vote on your answers as usual -- and they will be able to add their own answers. – Nick Cox Dec 14 '16 at 22:18
  • 2
    The example raises a different issue, which is that broad comparisons between the merits of different methods are sometimes considered to be too much a matter of opinion. But you can only try. The community has lots of ways of telling you quickly whether you have made a good start. (Then again people enjoy arguing about the messy parts of statistical science such as the importance of normality assumptions or what to do with outliers.) – Nick Cox Dec 14 '16 at 22:22
  • 4
    The open-ended nature of such a question suggests it would not fare well in that form on any SE site. – whuber Dec 14 '16 at 22:29
  • 4
    I think it's unfortunate that this is unlikely to work (remain open / be a viable question), but I think it's unlikely to work. – gung Dec 15 '16 at 3:20
  • The "Can anyone correct?" part works better than "Can anyone add to my understanding?", which is essentially inviting whatever people would like to add that they think might help. Too open-ended. But there's no hard and fast division between the two. – rolando2 Dec 24 '16 at 20:00
up vote 12 down vote accepted

In general specific questions are better than very broad open-ended questions (even when they're not too broad to stay open). "Add to my understanding" is not usually a good question.

In particular it's better to try to make a question that would be useful to the next person as well as yourself, even if a good answer ends up repeating some things you already know.

In addition, there's the problem that you might make your question nigh-unanswerable ("Yes, that's pretty much it" isn't really a suitable answer).

The main exception to this advice would be the special treatment we give to self-study questions (where describing what you already understand about the topic may be relevant).

You must log in to answer this question.

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