8
$\begingroup$

This recent SO blog post got me thinking about how DSP.SE can be more welcoming of newcomers of all flavors.

The specific sorts of questioners that I believe we need to encourage more are:

  • Homework questioners.
  • Too little information questioners.
  • Duplicate questioners.

What thoughts do you have?

For the first, I’m wondering if we should have a specific meta question with details about how to ask homework / self-study questions and links to good examples.


Interesting paper: Denae Ford, Kristina Lustig, Jeremy Banks, Chris Parnin, “We Don’t Do That Here”: How Collaborative Editing with Mentors Improves Engagement in Social Q&A Communities.


SO now has a draft Code of Conduct that is a little more extensive than their Be Nice policy. More details and reasoning behind it here and here.

$\endgroup$
  • 2
    $\begingroup$ And how to share related data in their purest shape (png images, csv samples) $\endgroup$ – Laurent Duval May 9 '18 at 21:06
  • 3
    $\begingroup$ Maybe we should have four meta questions: 1) How to ask good homework questions? 2) What information should you provide in your question? 3) Has your question already been answered? 4) How to add data (images, samples) to a question? $\endgroup$ – MBaz May 10 '18 at 15:31
  • 1
    $\begingroup$ We just need to add Image Processing explicitly. $\endgroup$ – Royi May 11 '18 at 14:14
4
$\begingroup$

i try to be nice to newbies and often modify their questions to demonstrate the proper use of $\LaTeX$. (it's really this ability to post equations and graphics that makes dsp.stackexchange.com displace comp.dsp.)

i have been kicked out of one SE site myself (just as i have been banned from Wikipedia, but i have been editing it anonymously since for more than a decade) and i think we should be inclusive of nearly anyone who is not themselves abusive or somehow trying to use this as an advertizing platform.

$\endgroup$
4
$\begingroup$

I think newcomers get intimated by the downvotes such questions do accumulate. A good way to go would be to try the be nice policy extensively covered in this meta question.

Regarding the questions, I do often add myself the "homework" tag to homework type of question, and explicitly mention to the OP to show their effort and where they got stuck. For questions with little information, it should be asked to provide more info on the questions. And for duplicates to simply link in comment where their question has been asked before and mark it as duplicate. There is also this quite detailed link on meta SE on duplicates. I think newcomers should get what's missing or what's not right in their question.

Another thing common with newcomers is cross-posting. This has been addressed quite extensively, cross-posting on multiple SE sites is not encouraged.

I fully endorse the idea of the meta question with links. As for the poorly formatted questions, I format as many as I can when I have the time and I think this MathJax Basic Tutorial and Quick Reference should be suggested in comments to such questions.

$\endgroup$
4
$\begingroup$

I am not sure I have answers to all of these and I have often thought about behaviours on both sides of the exchange, the person stating a question and the person trying to answer it.

Sometimes I feel that certain reactions are too abrupt. And it can be difficult to suppress that reaction you get when you are trying to understand what may be motivating such a "provoking" question. Some of them are obvious: "Do the research for me", "Write some code, I cannot be bothered", "Hey, what does this mean, because you know, I do have a connection to the internet but why use a search engine?....Thanks in advance...".

But the sieving is difficult because of language barriers, because of education gaps, even because of slight misconceptions.

So, we cannot but ask for clarification when in doubt. But, I am talking about plain, dry, question for clarification. Anything that shows irritation might be sending people away.

And it is not only the people who have registered but also those who are browsing and thinking about posting a question. We don't want to scare them away :)

EDIT:

Ouch...

enter image description here

From this question.

-----

As far as something a bit more specific to the topic of this discussion is concerned:

  • Homework newcomers

    • The obvious response here is "Show your progress" (but expressed less abruptly of course). But, what if they don't know where to start? I think that in that case, you need to start backtracking "To understand this, you need to have understood that and for that you also need to have grasped this one over here..." and so on. So, if they cannot show progress, can they at least provide basic definitions? The hints that are provided in the comments are a very good response to this, it's a relatively new thing, I do not remember hints a few years back. I think that those who care, will follow a knowledgeable dot trying to nudge them towards the right direction. Those that are in for the "quick reward" will simply give up. The problem there is, what do you do with that question which is not closed gracefully?
    • The only trouble I see with "...links to good examples..." is the copy-paste approach which optimises for wording...but not meaning :/
  • Too little information questioners.

    • This can be a disaster, from the point of view of the answer-er because it can get you off topic. So again...ask for more details. What I have been trying to ask in these cases is for the motivation "What makes you ask this?", "What are you trying to achieve?". This is the "slight misconception" trouble: "Oh, I thought I could...", yeah that might also be true but there is a much simpler way because....whatever. In this case, it might be a very quick "mind click" which then leaves us again with a sort of "Of course! How did I not think of this in the first place!" kind of closable question. What I have been trying to do in these cases is to remind / ask the OP to delete the question if they don't think it is "valid" any more.
  • Duplicate questioners

    • I really don't know about this. Why generate duplicates? "Ask once and you shall be migrated", some times incredibly fast.
$\endgroup$
3
$\begingroup$

I think we need to strike a delicate balance here:

  • be welcoming to people, because, hey, that's what we're here for
  • keep the platform functional (and that means sufficient-quality questions only)

I often go ahead and read a newbie question, and then pretty much immediately decide to close-vote it. That sounds incredibly harsh, but it's not: as new users, you don't even see the close vote count, AND, and that's more important here, I try to check back on the question within a few hours and see if things improved (close votes can easily be retracted).

Of course, improvement seldom happens spontaneously in this causality-centric reality, so I try to leave (hopefully helpful) comments. For example, a question that was

I would like to know the best audio-water marking technique with below operations,

  1. Add some audio-water marking in my WAV files (PCM samples) & store
  2. Do some processing like, equalization & volume/gain updating etc.
  3. Compare (1) & (2) to ensure integrity of audio content

Also what is the complexity & typical memory requirement (I need this for an embedded device)

would get a comment

Hi! Welcome here. I'd like to help you improve your first question for maximum effect. Let's take this apart: "The best" according to which metric? You can't say "the best car" without saying with respect to what (top speed,max load,fuel consumption,number of cup holders…). Then: Audio watermarking really is a broad topic and you can find very many, even beginner-friendly, resources using google, so we need to expect you to do a bit of your own research. My wild guess is that you've done quite some research (you did end up here, somehow,didn't you?). So, tell us what you've found out so far!

That's admittedly one of my explicitly more nice first comments.

Does this kind of thing work, where I do the following:

  • Identify shortcoming of new question
  • Close-Vote as appropriate per previous point
  • Leave Comment, encouraging improvement, often even in very simple terms (real-world analogies etc)

Often yes, but in the general case: not quite sure.

Of course, as said, people who are actively actually looking for solutions to a problem that they've really tried to investigate as well as they could are typically very positive about feedback and questions for clarification.

In the case above, the last two questions of OP were practically my indicator for "there likely won't be anything coming out of this".

We do have a lot of people that really try to avoid reading that book chapter and hope that we, upon being triggered by the right buzzwords, just magically funnel knowledge without effort of their own right into their cortex. I doubt that this typically is as consciously explicit as I just stated it! I instead assume that a lot of people simply are rather stressed when they come here, simply because signal processing is among the more math-y things they'll do and the work load of getting to grips with a chapter is very much having understood all the basic definitions leading up to that chapter; which more often than not means "basically all previous chapters" in their textbook. Since they don't (feel like they) have the time, they hope for shortcuts.

I think it's questionable we need to embrace these shortcut-seekers as is – simply because that leaves no positive impact on us, on the site (simplifying answers is what they seek, not correct ones), and questionable impact on them.

What we need to embrace is their potential to convert a "do my understanding for me" into a specific "I have a lack of understanding for this, how can I change that?".

So, what needs to happen is that we more clearly say

Hi! Welcome here. The way you ask this might mean you're having a hard time understanding your question's basics and hope we can give you a more concise answer than your literature research could do for you. We're afraid that's not the case. However, if you can rephrase your question to be like: "I need to solve {PROBLEM}, and for that I need to understand {ASPECT}" with the first ASPECT you encounter (don't be shy!) that you don't understand, you'll be getting forward in no time!

and really mean it. I don't object to answering easy, small questions, and if it is by citing definitions, if these are sufficiently researched.

The idea is that rewarding doing the decomposition of their problem into smaller questions that can individually be answered, maybe even by themselves, will lead to more questions asked after sufficient thought has been given, and the answers to these will encourage these askers to ask more, increasingly complex-but-well-researched (as opposed as complex and unresearched) questions.

That comes at two expenses:

  • If we need to encourage breaking down questions until each question asked occurs to even the asker as fundamental, then there's no "too easy" questions anymore.
    I'm guilty of thinking those exist myself (example comment). Even if its just asking for a wikipedia'ble definition, we can't just downvote and move on, but need to feed back "hey, did you try looking for this on wikipedia?", and only after that downvote.
  • We need to really be more clear with what we "demand". (we offer free services, and recommend people do something so that we can, but it still feels like a demand to many of those asking)
    That means actually writing

    Hey, I know this might seem harsh, but I had to downvote your question because it just dumps a complete homework problem on us, without even trying to identify what's hard for you, and what you can do; that really gives us nothing we can answer without writing a whole book. Give us the first thing that you need help with, and we'll try to help you. (Please do so by specifying that as a question by editing your post.)

I think that the frankness here is a feature. There will be users who get cross at us for "refusing to do our job", but let's be honest here: We don't have to help everyone, and our (as in: important answerers on this site) mental state does influence our willingness and ultimately our ability to help.

$\endgroup$
2
$\begingroup$

I think that someone new to DSP will generally have problems formulating specific questions. You have know something to understand what you don't know. Some people will ask a vague question. Some people ask shot gun questions. Some people don't really know why they are asking.

It can be hard to guess what they should be asking.

As far as homework questions go. We really don't know if a problem is assigned as a collaborative question or a strictly individual one. We don't know if the question is a take home test question. Cheating is a big problem at my school. I don't think anything beyond a hint should be offered, and only if they show effort.

$\endgroup$
2
$\begingroup$

How about being welcoming to new people that answer questions?

I’ve tried several times to become active here, but:

  • The front page is filled with old questions bumped by Community because nobody cared to upvote the answer. (Community bumps questions that have answers with 0 votes and not accepted answer).

  • Answers written with lots of effort and detail get no attention whatsoever. I’m left wondering, does anybody read the answers? What is the motivation of writing an answer if nobody reads it?

  • Some of your log-time established users seem a bit... protective of their turf? I’ve seen some pretty aggressive responses to answers.

If you want to be welcoming to new people (both answers and answerers), vote more. Vote often and vote hard. Use up your 30 votes every day. Even negative votes are more welcoming than no votes (at least if they come with an explanation).

Here’s site stats on badges given — 5.3k users have upvoted, only 45 have upvoted 30 times in one day. Granted, it’s hard to vote 30 times in a day. But that is exactly my point: you’re not going to get to your limit, there is no reason to hold back votes.

enter image description here

$\endgroup$
  • 2
    $\begingroup$ We really need to take care of those questions. Many times the answers are great yet the OP is either not ware of the system to mark them and upvote them or just left the site. We, all of us, need to be generous and mark answers up when they are good. We should do that more often. $\endgroup$ – Royi Aug 21 '18 at 7:41

You must log in to answer this question.

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