How can we encourage (new) authors to ask confident questions?

In Clippy's voice:

voo-oop, I see you [are new] | [have a terrible track record of asking questions], would you like to see some suggestions for newcomers?

Suggestions for newcomers section:

This site is not some forum which takes ages to respond so stay active.

Everyone's time is valuable so do not make people pull teeth to get information from you.

Everyone is a volunteer so do not demand answers but do slam your fists on your desk and demand pictures of Spider-Man.

Before asking:

• Google your general problem and see if there are solutions that you can try
• If you have an error message or number then that is usually a great start for Google
• Keep track of the solutions you tried and why they failed

Asking:

• Try to provide the minimal code necessary to reproduce your problem and make sure the problem exists in the minimal code provided. People will ask for more code if need be.

After asking:

• Make sure to do absolutely nothing but watch your question for the next 10 minutes and promptly answer any questions which will inevitably arise. This site receives x number of questions per minute for your [most-popular-tag] tag so your question can quickly be ignored and lost.
• If asked for debugging information, edit it into the question; do not comment!
• If asked for debugging information then provide it immediately or provide a comment stating that you will have it in x number of minutes.

Growing up in a very polite and gentle family, I found it natural to preface questions in class with, "May I ask a question?". Some teachers would respond, "Yes, and you've just asked it." I felt embarrassed; why did they have to be so rude? Why couldn't they just say yes and let me ask my brilliant question? Slowly I realized my teachers were showing me that I was the one wasting time -- I should just ask my question.

My early confusion came from a conflict in cultures perhaps not dissimilar to they one some new-flowers face when asking questions to us overflowers. But I believe we tend to learn to adapt to the culture that we are in. So let's try to welcome / acculturate new-flowers by trying to be respectful and patient while showing by example how people are expected to behave on Stack Overflow.

Instead of telling people not to apologize, beg or be excessively deferential, let's simply edit it out. If we show what a good question looks like and if we can give a good answer, they will hopefully see the merit of our way and eventually learn how to ask good questions on their own (and, hopefully, contribute good answers too).

You seem to be looking for a way to change peoples' behavior (either the askers' or the answers' or maybe both). I'm of the opinion that telling people what to do (e.g. through pop-ups) is generally not an effective way to change behavior. (The help links have lots of useful information on how to ask a good question that is ignored by thousands every day :)).

This podcast describes some research you may find interesting. It suggests that while telling people what to do is ineffective, instead changing their perception of cultural norms can be effective. I believe, applied here, that means the best we can do is lead by example. Do the hard work of editing out blather one question at a time, showing through action the social norms which are expected.

Babies and animals are quite good at picking up on tacit patterns of behavior. I have hope we adults haven't lost this ability either, and with a little time on the site most will come to understand we want clear, concise, answerable questions without the blather. For those who need a bit more concrete explanation, I hope a link to this meta question would suffice.

Here is an attempt at quantifying the effect of using the word "downvote" in a Stack Overflow question. (I'm using the presence of the string "downvote" as an indicator for questions that are likely to contain some variation on the phrase "Please don't downvote me".) This SEDE query finds the average score received by first-time question askers if their question body contains the string "downvote":

SELECT sum(p.Score) as total, count(p.Score) as num_questions, (sum(p.Score)*1.0)/count(p.Score) as ave
FROM POSTS p
LEFT JOIN USERS u
ON p.OwnerUserId = u.Id
LEFT JOIN (
    SELECT u.Id as u_id, min(p.ID) as min_pid
    FROM POSTS p
    LEFT JOIN USERS u
    ON p.OwnerUserId = u.Id
    GROUP BY u.Id) as mp
ON mp.u_id = u.Id AND mp.min_pid = p.Id
WHERE p.PostTypeId = 1 AND p.Body LIKE '%downvote%' AND p.Id > ##cutoff:int##

By varying the cutoff parameter, we can see how the average score changes over time. (By decreasing cutoff, more posts are included in the query and since the post Ids are assigned in increasing order, a low cutoff value corresponds to a SEDE query which includes older questions.) The cutoff was also necessary since without the cutoff the SEDE query times-out -- the query takes too much time to run on the full database.

Here is the raw result:

|   cutoff |     sum |   count |     ave | contains "downvote" | last n posts |
|----------+---------+---------+---------+---------------------+--------------|
| 50000000 |     -23 |      25 | -0.9200 | yes                 |        83751 |
| 48750000 |     -83 |     195 | -0.4256 | yes                 |      1333751 |
| 46250000 |    -174 |     515 | -0.3379 | yes                 |      3833751 |
| 45000000 |    -188 |     681 | -0.2761 | yes                 |      5083751 |
| 40000000 |    -216 |    1224 | -0.1765 | yes                 |     10083751 |
| 50000000 |   -3438 |   37008 | -0.0929 | no                  |        83751 |
| 48750000 |   82312 |  520665 |  0.1581 | no                  |      1333751 |
| 46250000 |  386292 | 1443936 |  0.2675 | no                  |      3833751 |
| 45000000 |  583468 | 1901119 |  0.3069 | no                  |      5083751 |
| 40000000 | 1549510 | 3589199 |  0.4317 | no                  |     10083751 |

Since the maximum Post.Id is currently 50083751, we can compute the number of posts searched in the query by defining last n posts = 50083751 - cutoff.

And here is a graph of the average score versus last n posts segregated by the presence of the word "downvote":

This seems to show that including the word "downvote" in a question earns the asker about -0.9 votes initially, and their question might be worth about +0.6 more votes (=0.4317+0.1765) in the long run if the word "downvote" were removed.

While this phenomenon certainly exists, I don't think it is unique to SO and I don't think some elitism reputation has much to do with it.

As you observe, there's a lot of cultural aspects here and that's probably one of the main reasons. Strange translations to English is another.

Also, a whole lot of people asking are students, who may be used to address their teacher in a certain way, often respectfully. Then they address the SO community as they would address their teacher. And how they do that can in turn can also be a cultural thing. There's for example the incredibly common "I have a doubt" almost exclusively coming from Indian students (see this), which sounds strange to others but simply means "I have a question".

"I'm new to this / programming language / SO"

might actually be a valuable piece of information. You may realize that the person asking the question may not know about some of the rather obvious ways of obtaining the answer to the problem.

You may e.g. realize that while the answer to a question is already given somewhere, a "newbee" would not be able to find this answer. Or to understand it. It can make the difference between downvoting for no research shown or leaving a nice comment about the documentation using italic font for particular methods in use or such things. When closing a question as duplicate I also found myself leaving a comment explaining in how far this is a duplicate more often in cases where I realized the questioner couldn't have found this duplicate from where they started.

More generally, the very same question is really a different one being asked by some expert compared to some inexperienced person. Consider the question "Why does water change its color in the evening?"; imagine this being asked by a child and by a physics professor. It may in both cases be a valid question - with no research shown, because the child doesn't know what to ask previously to that to acquire more insight necessary to solve this, and the physics professor may simply assume that you know in how far this has, despite everything being written about it already, not an obvious solution. If anything, you still need to condemn the professor for asking a bad question, not the kid.

So maybe this particular phrase does not actually qualify at all for being removed.

Similarly,

I am stuck for days
I have googled and read every single post on stackoverflow

assuming that those statements are actually true (which, agreed, is often not the case), they could be translated into "I have an XY problem, but I don't know how to phrase X, so I present Y here in hope someone can point me to X". Those are then also valid extra information to consider when writing an answer or asking for clarifications in a comment.

Possible actions to take:

I'm a big fan of some interactive guide to asking questions. Hence If the user types in any of the meantionned phrases yellow popups could occur to inform them about the consequences.

• "I'm new to ... -> "While you may have the wish to tell people about you being new to ..., make sure that this is actually relevant for the question to ask. In most cases what you rather need to inform readers of your question about is what exactly your level of knowledge concerning ... is. State which resources you have used in the past, which tools you are aware of and confident in using, which part of the problem you do understand how to solve and guide the reader to the problem by clearly stating what piece of information you lack to solve it yourself."

• I am stuck for {days, years, ages}. -> "Be aware that the amount of time you are stuck is not relevant to the problem. Instead tell the readers of your question what you have attempted to solve the problem and which resources you have been using."

• I have googled... -> "Make sure not to forget telling the reader about the result of your searches and readings. Also link to relevant resources and questions that have not helped you and state in how far they don't solve your problem."

As a new poster on Stack Overflow [meta]: I do not fear the asking. I rather fear the anger towards the posted question. If you have a high ranking, you're automatically regarded as asking only the necessary. If you have only a few posts, your question is surely poorly researched and gets downvoted in seconds.

//e1: This is not meant as a general statement concerning every single poster, but it was rather thought of as a starting point in the theory as why one might have a (psychological) difficulty asking a question. It is not to be regarded as a 'verdict'.

//e2: @halfer: I constrained myself to the actual question, not the other theories stated thereupon.

I agree with unutbu and ImportanceOfBeingErnest but I don't have sufficient reputation to comment or upvote them, hence my answer. Is it ironic that I provide that qualification? Newer users are constantly reminded of their new status by the hierarchy of badges/reputation points which allow or prevent certain participation. I understand why they exist and do support them, but the constant reminder may be a contributer to what is described in OP's question.

Re. unutbu: I had my first question heavily edited and I learnt from it. There are so many rules/guidelines that I was not able to process them all when I posted my first question but through the example I was shown, I learnt. My personal experience aside, modelling correct behaviour is a most effective tool and definitely applies in this situation.

Re. ImportanceOfBeingErnest: I would also write

"I'm new to this"

in a question, but not as some type of grovelling. My aim would be to advise any prospective helpers that I would appreciate a response given in the simplest of terms since a complicated and in-depth response may be beyond my current level of understanding.