11

I am a voracious StackExchange user, always I have a technical question I go to StackOverflow, also every morning I like to come here to programmer to read some more advanced questions and opinions about design stuff and programmer challenges (as opposed to programming challenges). I noted, a long ago already, that we have a great number of negative voted question. I can understand the motives. 1 2

But why can't we raise up the quality of the Questions? Look at this chart I extracted from data.stackechange. It shows the ratio of down voted questions to total questions asked that year.

enter image description here

Despite its huge size, StackOverflow could, given the years control (and for some time, lower) the negative voted Questions. In Programmers in other-hand negative question have increased over the years (except in 2013-2014 period, where it appear to be controlled).

  1. What can we do to help with the questions quality?
  2. Are we already controlling better new posts in the review queue?
10
  • 9
    There is a huge confounding factor in your analysis, and that is that SEDE doesn't contain deleted questions. Commented Nov 13, 2014 at 11:27
  • 4
    as @MadScientist points out, SEDE can't account for the fact that all posts with a score of zero, no answer, at most one comment and low views are deleted after a year. That makes it less than useless (=harmful) to analyze issues like you try to investigate; a feature request to fix this is Can some metadata about deleted posts be included in Data.SE?...
    – gnat
    Commented Nov 13, 2014 at 11:45
  • ...meanwhile, the only option is to request SE developers to give a correct aggregate data for this query, that is data that takes into account deleted posts
    – gnat
    Commented Nov 13, 2014 at 11:47
  • Another explanation for your data would be that the users on Programmers are more prone to downvoting. ;-) Commented Nov 15, 2014 at 19:27
  • 10
    I've been saying for a long time that the site needs a name change, and I still believe that. The site is called "Programmers", but is not about programmers. It's no wonder that people get confused and post questions for or about Programmers, even though they are off-topic for the site.
    – Rachel
    Commented Nov 17, 2014 at 16:21
  • 3
    @Rachel: I would favor renaming the site "The Whiteboard," and renaming the chat room to "The Blackboard." Commented Nov 21, 2014 at 23:20
  • @RobertHarvey I would be in favor of that too. Coming to a Q&A site named "The Whiteboard" gives the expectation that it's a Q&A about whiteboard-y design questions, unlike "Programmers" which gives the first impression of being a site to ask other Programmers anything.
    – Rachel
    Commented Nov 23, 2014 at 20:02
  • Um unless I am mistaken, I think you have a huge flaw in your Data.SE query. It's looking at Answers (PostType = 2), not Questions (PostType = 1)
    – Rachel
    Commented Jan 27, 2015 at 16:46
  • If you want a look at the ratio of negative questions asked per year, this query is more accurate, but keep in mind it does not included deleted posts, and there is an automatic script which deletes a large number of low scoring posts.
    – Rachel
    Commented Jan 27, 2015 at 16:54
  • see also at MSE: Ways to give users some specific education about question quality and topicality
    – gnat
    Commented Apr 20, 2016 at 10:08

4 Answers 4

12

Part of it is trash questions, but there is also questions that are not in scope.

Right now the scope is detailed in Help -> Help Center -> What topics can I ask here?

Is it there? Yes. Is it good enough? Not really. If it were.. well then gnat's life wouldn't be as hard as it is.

That really isn't programmer's fault, thats just how all of SE is. It starts from position of assertion that you should read a help-center first. The fact that nobody does this hasn't been embraced yet.

Users learn what each particular site's unique interpretation of what is 'subjective' and what is off-topic by visiting one of the on-hold questions, or having one of their own questions put on hold.

So education usually comes after the mistake. SE might want to think about how to improve their UX in order to educate during question composition, instead of burying the scope three clicks deep or having to be linked to after a question is put on hold.

And yes I know this is already done when the cursor's focus is in the question title, but as we can see, that obviously isn't good enough.

5
  • 2
    I've been meaning to ask this question for a while now. When I read questions at night I often see a high amount that are downvoted - I've always been curious as to why. I think the answers here are good, in any case. Commented Nov 18, 2014 at 4:05
  • Another idea is to award points for taking an off-topic or subjective question, and if it is possible to derive an underlying on-topic, non-'subjective' version of the question, and help the new user ask the question. Its a harsh first impression of the SE network to ask an honest question only to be put on hold and downvoted 10 times. I personally think a simple downvote/upvote isn't very good way of communicating. There are multiple reasons for upvoting and multiple reasons for downvoting. The oversimplified up/down isn't sophisticated enough for such a sophisticated community imo. Commented Nov 18, 2014 at 15:49
  • A downvote is sometimes used to agree with the decision to put a question on hold as off-topic or subjective. Theres a better way to communicate that than a downvote. A downvote is interpreted negatively by the poster. Its especially harsh when asked to a very active community, where the person asking the question can be dogpiled with downvotes. Clearly communicating that 10 people agreed that the question is off-topic would be a much better impression than just a bunch of downvotes. I think SE can do a lot better than "you didn't read the help-center, so take that and go away noob". Commented Nov 18, 2014 at 15:58
  • And if the new user can find a way to re-shape their question with the help of others, reset the downvotes and award points as sort of a tutorial achievement. Would be much more inviting. Commented Nov 18, 2014 at 16:03
8

We are inline with similar sized sites

There's an interesting chart there... but I draw issue at comparing P.SE to Stack Overflow. In particular, the size of the voting user base compared to the daily influx of questions can make most of the stuff on Stack Overflow get no votes at all (there's also a significant 'whine about down votes' and 'give sympathy up votes' (examples 1, 2, 3 (entirety of question in mouseover in case they are deleted)) culture on SO).

I wanted to see what similar sites to P.SE looked like - similar in terms of questions/day and overall size. This is what I came up with:

Other sites

You will see that we are by no means the most down voting of the 30k or so question sites that I sampled (Gaming, Programmers, Electrical Engineering, Android Enthusiasts).

The question really is "what is EE doing right?" along with "what groups Gaming and AE together?"

The answer to this is likely the accessibility of the material to the 'common' person. The harder/less accessible the material, the more focused the site can be and clearly communicate that focus to people asking questions. Furthermore, the more difficult the material, the more natural it is for the person to have done the requite study of the problem before asking the question along with asking the question fully.

Judging from this, I would suggest that if we want to have fewer down voted questions, it is would be necessary to restrict our scope drastically such that we're closer to that of Theoretical CS or Server Fault in what is allowed or not and take fewer soft questions; requiring all the questions to be at the professional software architect level.

I'm not sure that would be the right thing to do.

Furthermore, there's the conception on other sites (Stack Overflow, I'm looking at you) that Programmers.SE is for questions without code or for getting opinions. One can reliably go and find belongs on P.SE type comments on Stack Overflow (data.SE) for questions that don't meet the desired quality. The questions get asked here, and down voted, and, well, that's what you see. I contend that it is impossible for P.SE to stop the problem as there are far too many with this misunderstanding of P.SE's scope on SO and the best we can do is close, down vote, and educate people who post here under those conceptions (and, again, that's a non-negligible portion of the ratio).


Changing the name won't help

One reoccurring theme that has popped up over the years is "maybe if we change the site's name people will post on topic questions?"

I'm pulling up the current newest questions page and listing all the questions with a negative score:

It is my contention that absolutely none of these would have been not asked if the site's name was SoftwareDevelopment.SE or SoftwareEngineering.SE. Every single question there would still have been asked. Very few of them had off topic votes. General topicality is not the problem with the questions (specific bits of 'tool recommendation' and 'career advice' still is).

We are long past the days of "what is your favorite comic" or "what should I name my cat" and the like. Those questions don't show up. Programmers is understood to be a site about software development processes.

The off topic questions that we get now are ones that very closely map to the topic of software engineering, but are issues that we are not qualified (as programmers) to answer. Career advice and legal being the two biggest ones. We also regularly get polling questions, but that isn't an issue with the site name, that is a misunderstanding of the Q&A format itself (and a failure to read the tour which shows what is on and off topic).

But there will always be people that suffer from the "herp derp oh look a textbox" syndrome, and no amount of renaming the site or amending the FAQ is going to fix that. -- Anna Lear

5
  • Well, in my opinion it's a lot easier to ask off-topic questions on Programmers.SE than a lot of other sites (like EE.SE or AE.SE)... There are so many computer science-related SE sites that I would say, in general, it is easier to ask off-topic questions here, for example. Subjectivity is another thing, but the idea probably presents itself across the various reasons for why people downvote questions. Commented Nov 18, 2014 at 4:03
  • 1
    @ChrisCirefice with the exception of career advice, legal advice, and recommend a tool... we don't really close a lot of questions as off topic. On the other hand, we get a lot of too broad, unclear, and should have been asked on Stack Overflow (with the caveat that we really try to avoid migrating crap and tend to keep a number of those questions here than migrate). Granted, stats would always be a good thing here and I don't have them readily available. However, all of this is fairly clearly shown in the tour and the help center... but getting people to read those is another challenge.
    – user40980
    Commented Nov 18, 2014 at 4:51
  • 1
    To be honest, I've only ever read the help center for English Language and usage, because I wasn't sure if my question would have been off-topic. I usually just peruse the content of SE sites and gain an understanding of what types of questions should be asked. I've only had one off-topic question, and that was here on P.SE. For new site members, it makes sense to read the help/tour, but I'm a perfect example of that not happening, and I frequent SE sites regularly throughout the day... Commented Nov 18, 2014 at 5:32
  • 3
    @ChrisCirefice in days of old, the general thing to do with a new community was to lurk for awhile. The people that jump in are the ones that should/need to read the faqs. That you lurked first to gain an understanding is completely acceptable and encouraged.
    – user40980
    Commented Nov 18, 2014 at 6:17
  • The graph in the question is incorrect, so if this graph is based on the same data, it is also incorrect. Can you update it so it's based on questions instead of answers and see if it is the same? If you want an easier to use data.se query, try this one
    – Rachel
    Commented Jan 27, 2015 at 16:56
7

Your data is inaccurate because SEDE doesn't account for deleted questions but anyway, there are few factors worth thinking of.

In the past, Programmers were getting a reasonable mix of good and bad questions related to careers, software recommendations and code reviews. Even when topicality of these was slippery, well presented questions about interesting, real problems were considered worthy of upvotes by many readers.

Nowadays, we lost a good part but bad part remains with us.

I mean, users capable of asking good questions in above topics, eventually find their way and stick to specialized sites - TWP, CR, SR.

  • What is left to us are askers who want to ask but can't make a good question. They post (or cross-post) their stuff at Programmers, Stack Overflow, Yahoo Answers, anywhere, in a hope to find a kind soul to feed them. This contributes to lowering average quality of questions in these topics.

Another stream of low quality questions are those from users blocked / warned at Stack Overflow. Per recent stats, we're getting 4-5 delete-worthy questions a day from these folks.

In the past, they could pollute SO with their low quality stuff for months or even years but for last year or two, due to improvements in burning down SO close queue, they started hitting question blocks and warnings more and more.

  • Per stats, Programmers appear to be most popular of SE sites among this category of askers. After being blocked or scared by warnings at Stack Overflow they simply try to dump their low quality stuff over here, thus also lowering average quality level.

If honestly, I can't see what we can do to help with the quality of questions like above, nor how we can control these in review queue.

We can't realistically expect askers of bad questions to "refocus" to specialized sites because at these sites, such questions seem to be detected and rejected even faster than over here. This is probably because of clearer, narrower specialization regulars at these sites build a better sense of what is good and what is not (at least that's how I feel at TWP).

I also can't imagine how it would be possible to make users blocked / warned at SO stop their attempts to leak through Programmers. I can't think of a way to tell someone facing a block message at SO that they shouldn't try their luck at Programmers.

Wait! You've recently been blocked (warned) at Stack Overflow. Consider that quality requirements at Programmers are similar. Note that Stack Overflow topics are not covered here and that your questions can't even be migrated to Stack Overflow because of the block.

Are you sure that you want to proceed asking question?

Click No to cancel asking.

Click Back to return to Stack Overflow.

-1

Update for current page:

A look on the page of the first 50 newest questions counts:

24 questions are downvoted 13 questions are upvoted 13 neutral

This is a ratio almost 2:1 down/upvote

7
  • 1
    Eh? We get ~35/40 questions per day, your stats will be pointless tomorrow.
    – yannis Mod
    Commented Mar 18, 2015 at 10:23
  • I don't understand your point. of course they will be. I am raising the point that a majority of the questions are being downvoted, agreeing with the topic here and making it more obvious with an example because as previously said, original comment doesn't include deleted topics.
    – arisalexis
    Commented Mar 18, 2015 at 11:04
  • 1
    All your data show is that the majority of incoming questions today were not well received. Why is that a cause for concern, or even interesting? We have good days, and we have bad days. Given that our average is ~15% downvoted I think it's fair to say that today is just a bad day.
    – yannis Mod
    Commented Mar 18, 2015 at 11:20
  • where do you get this 15% number? does it include deleted questions? also the ratio down/up is much more interesting than the raw number given unvoted questions are way too many.
    – arisalexis
    Commented Mar 18, 2015 at 11:23
  • 1
    data.stackexchange.com - it doesn't include deleted posts
    – yannis Mod
    Commented Mar 18, 2015 at 11:26
  • since probably most of the downvoted posts are deleted and almost none of the upvoted, it is safe to assume 15% is very far from the actual number. Now again, this is why I wrote my comment and you can see it day after day that "today is not an unlucky bad day" and raise awareness.
    – arisalexis
    Commented Mar 18, 2015 at 13:15
  • @YannisRizos FWIW per my observations over last few months, there seems to be more low quality questions posted on Tuesdays and Wednesdays. Would be interesting to see if this correlates to stats on users discovering that they hit question block / warning at SO (as far as I know SE team keeps track on blocked attempts for spam but I doubt that they do this for attempts at asking stopped by block / warning)
    – gnat
    Commented Mar 18, 2015 at 16:30

You must log in to answer this question.

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