31

I was scrolling through the list of newest asked questions and noticed that only 3 out of 15 had non-negative votes for the question. Is this just an anomaly or is the quality of the average question being posted so poor that only 20% deserve a non-negative scoring?

I would think that with such a low success rate, many first time posters would end up being turned off by the site and never reposting if they have such a low rate of posting a successful question.

What is the historical rate of site visits and how is this trending over time? If it is going downward, that would not bode well for the future of this site, perhaps such down voting is a contributing factor to any such possible pattern? Could the down-voting be handed out more conservatively and instead some of these questions be edited to make them worthy?

  • 6
    just a lot of crappy questions in the past day? It's not always like that. – Mitch Aug 24 '12 at 20:27
  • 13
    Agreed! I have only been on ELU for a couple of days, and I am very sad to see so many downvotes on both questions and answers, especially without any explanation to aid improvement. I thought the point of downvoting was to improve the site. The overall feel of ELU to me is overly negative and unproductive. It is very different from my experiences on other SE sites, where you only downvote for very poor behavior or grossly inaccurate answers and there is a spirit of understanding and cooperation. – Rachel Aug 24 '12 at 21:25
  • 5
    It is not so different from other sites: it exists to serve experts. The trouble is that this experts site is being overrun by very basic questions, because there is a real need out there for answers to basic questions. As I mentioned in a comment on another thread, please support the English Language Learners proposal on Area 51, which should help both communities. – MetaEd Aug 24 '12 at 22:31
  • 6
    @ΜετάEd That is not a correct definition of SE sites. They are not meant for experts only. – Alenanno Aug 24 '12 at 22:48
  • 14
    @Alenanno “Stack Exchange … communities [are] dedicated to serving experts in a specific field”. (About Stack Exchange) “This is a … Q&A for linguists, etymologists, and serious English language enthusiasts”. (About English Language & Usage) If SO tech sites were being overrun with questions about whether it is right to use a comma or a period after a statement in C, or how to spell “printf”, we would see there exactly what we see here. This is why the ELL site is sorely needed. – MetaEd Aug 24 '12 at 23:36
  • 2
    @ΜετάEd Some lines after what you quoted: "our communities are built by and for those best able to define them: the experts and enthusiasts" (About Stack Exchange) See? It's not only for experts. They are needed, but they are needed as much as the less experts. P.S. Please consider making your first letter a Latin M, so I can quickly answer you without manually copying your nickname. :) – Alenanno Aug 25 '12 at 9:11
  • 8
    As a long-time contributor ELU, I have to agree with this sentiment: I feel that our site is becoming less and less welcoming to new users. It was very different in late 2010; the site has changed gradually since then, and not in a direction that I like. But the majority is just like that, and there is nothing the minority like us can do about. We must concede defeat. – Cerberus_Reinstate_Monica Aug 25 '12 at 15:44
  • lol @FeralOink How the heck did you find it? "Cryptography with no algorithms" is kind of a stretch though :) – Adnan Aug 27 '12 at 13:28
  • @Adnan It was very easy! You are the computer security expert, I am merely Ellie! Take a look on your Pseudo-Random blog, most recent post, okay? Your thesis said there was no use of fancy cryptographic algorithms. I confess, I didn't take the time to read it though, it is 108 pgs long ;o) – Ellie Kesselman Aug 27 '12 at 18:07
  • 1
    As a new user I find it a bit disconcerting as well. And half the time I'm not sure why topics get downvoted and/or closed either. Just earlier this week there was a good question about related compound words (fernweh, nostalgia, wanderlust). As a historical linguist I can see exactly what the question was asking, but as was the case with this one it was closed before I could respond. I think for a start questions should not be closed at all unless they are completely off-topic/vulgar. At least give a few days for someone to respond; we don't all sit on this site all day long. – Richard Burian May 31 '15 at 4:34
  • 1
    I'll also add that I noticed several questions downvoted when the person posing the question was obviously not a native speaker. Does one have to have flawless English to be a linguistics enthusiast? They teach linguistics all over the world and they teach English all over the world, too. It's natural to assume that foreign language learners would have linguistics questions since this site is probably the only one they will expect to get answers from; and a couple of universities I know of in Europe even recommend using this site to get answers. – Richard Burian May 31 '15 at 4:40
25

You have made a very good observation. You may recall that the last time we communicated, almost a year ago, I told you that I thought that the silent downvotes on your question were due to "overzealous enforcement of the FAQ guideline that suggests that off-topic questions should be down-voted to remove them from the front page, and that answers to off-topic questions should be down-voted to discourage answering off-topic questions."

In general, I think this explains the accumulation of downvotes on questions that might otherwise be simply closed. However, there has been a recent trend of increased downvoting in the past week. Prior to that, the level of downvoting was pretty consistent for months, and most closed questions were downvoted only enough to remove them from the front page (four votes).

The mods have found this concerning and we have been discussing how to address the issue.

The problem is that either we have a front page full of closed questions, or we have a host of downvoted users who may just need help asking their question, and either way it makes us seem unwelcoming.

Additionally, we have new review tools and many of the 10K+ users have spontaneously started a site cleanup. I think the excitement has created an environment of hyper-awareness to topicality and potential question duplication that has translated to a large user pool thinking that they should downvote every question that might have been asked before or that might be off-topic.

That said, let me take this opportunity to remind the community that downvotes are really for egregious behavior or poor answers. Be mindful of a user's rep and how long they have been a member on the site before you downvote or leave a snarky comment.

For our voting community

Before you downvote:

  1. Did you ask the OP to clarify and give them at least one day to do so?
  2. Did you edit the OP's question to include information that the OP had buried in the comments?
  3. Did you clean up and retag the question so it was clearer and easier for other users to find (even if it was a dupe and even if it was off-topic)?
  4. Is the OP a new user who needs some extra consideration because maybe they can't figure out how to edit their own question, or maybe they've been overwhelmed by an overgrown comment chain?

You are not required to leave a comment if you downvote, but the site won't improve if you don't.

Downvoting should be the last thing you do, not the first.

  • 3
    This is a very informative answer, and its substance - especially the "For our ... " part - should be prominently posted. As a very new user myself, I find this sort of practical exposition of site etiquette far more useful than abstract principles. – StoneyB on hiatus Aug 25 '12 at 0:41
  • 2
    I want to add that most of the time, a duplicate question should not be downvoted: if the original was appropriate to the site, then by definition the duplicate is also appropriate. Save your downvotes for the truly egregious "research? what's that?" questions. – Marthaª Aug 25 '12 at 4:21
  • 1
    "Downvoting should be the last thing you do..." So downvoting should come after voting to close or flagging? – Hugo Aug 25 '12 at 4:54
  • 1
    And what about those very simple questions that are answered by the most basic Google search or dictionary check but show no research whatsoever? – Hugo Aug 25 '12 at 4:57
  • @Hugo I think she meant that before downvoting you could try to correct the wrong part, comment explaining or asking the author to fix something, etc... This means to be involved in the community. I usually downvote when posts really deserve it. In other cases I comment and/or edit to fix what was wrong. Sometimes I vote to close without downvoting, because an off topic question can still be a great question. – Alenanno Aug 25 '12 at 9:19
  • 2
    @Hugo: Google is not general reference. – Marthaª Aug 25 '12 at 14:30
  • 1
    @Marthaª: No, but Google shows research effort, and would render some simple questions unnecessary. – Hugo Aug 25 '12 at 15:05
  • 3
    I've contributed to the upsurge: I started downvoting QQ about a week after I came on, largely because that "You haven't voted on a question in a while" popup made me feel I was supposed to downvote (and upvote) more. Perhaps the popup should be reviewed. – StoneyB on hiatus Aug 25 '12 at 16:53
  • 1
    I have to disagree that downvotes are for egregious things only. The tooltip sets the bar simply at "not useful", and quite frequently I see folks saying "don't take downvotes personally" and this question even seems to encourage downvoting when you vote to close. – Lynn Aug 25 '12 at 20:50
  • @Lynn The answers on that referenced question specifically say that you should not downvote when you vote to close unless the question merits a downvote for other reasons. Being a duplicate or off-topic is not sufficient cause for a downvote. – Kit Z. Fox Aug 25 '12 at 21:26
  • 1
    @KitFox: Yes, but the question itself got +5 votes, though, which (I thought?) indicated a general agreement? Also, a lot of our closed questions are general reference/show little research, which is mentioned in the comments as reason enough to downvote them. – Lynn Aug 25 '12 at 21:53
  • @StoneyB I have had that thought about the pop-up. I don't recall if it just does it on questions or on answers too. I don't think it is necessarily bad, nor that it should be changed, but merely clicking up your comment was insufficient to express my vigorous nod of "yes, I so know what you mean!" – Ellie Kesselman Aug 27 '12 at 6:18
  • I would like to suggest another point: "5. If the question does not fit EL&U, does it fit another StackExchange site, for example ELL (when it has been committed)? Then migrate that question (or flag for migration) to the other site." – Stephen Oct 14 '12 at 14:01
14

I can remember having a distinct thought yesterday, about how there seemed to be a dearth of good questions, and a rash of some very bad ones.

("Good" and "bad" are subjective, so, to save time, you can read my thoughts about "good" questions by looking at my profile; I've already written about it there.)

I don't believe this is a matter of experts vs. enthusiasts, or EL&U being a "negative" community. I'll pick one downvoted question from yesterday (I picked this one at random; it was the first one I picked, and it will serve my purposes just fine):

enter image description here

Here are the thoughts that run through my mind as I read this:

  1. Yet another which-one-of-these-prepositions-should-I-use questions? Sigh.
  2. Why is this person putting a space before each question mark?
  3. "Or, each one gives a different meaning to the sentence?" is VERY bad grammar1.
  4. Did this person even bother to look up these prepositions in a dictionary? It doesn't look like it to me.
  5. Where did that quote come from? Is it made up? Or pulled from an employee handbook?
  6. What does that quote even mean? I don't exist until I'm absent twice?
  7. Why is the O.P. "in doubt" about this? Is it really important?
  8. Does this question, in its current form, make EL&U a more interesting and informative place for the community at large?
  9. How hard did this person try to research and answer this question before they posted it here?

As for that last question, I've conveyed that sentiment on meta before, and some people have taken umbrage at that thought. I'll say this: I work hard on my answers. I perform research as I write them (I just double-checked the meaning of umbrage in a dictionary, for example). I proofread, edit, and spell-check. I do searches on the web to validate my hunches. And I don't think it's unreasonable to expect others to expend a similar level of effort into crafting their questions. (If you want to ask something trivial, and expend minimal effort, then I highly recommend Yahoo! Answers, where there is no downvote button!)

I still remember when I first read the FAQ six months ago. The FAQ seemed serious, almost grave, and I was intimidated – I did not want to come on here and make a fool out of myself.
I wish some other users would feel that same sense of gravity before they started typing out a question.

The O.P. of this meta question said:

I noticed that only 3 out of 15 had non-negative votes for the question.

Okay, maybe that looks bad on the surface. But how many of those 12 negatively-voted questions had edits by their authors? How many people took that feedback as an impetus to improve their question?

Moreover, how many of those negatively-voted questions were about basic grammar? How many bothered to cite a dictionary? How many provided context, to show how the question might be based on – per the FAQ – some signficant, real-world problem? How many were about a problem that would interest a serious linguist? Looking at some negatively-voted questions from yesterday, I find:

  • "An alleged burglar was left with a broken nose, jaw and ribs": should the sentence have been framed better to convey that the nose, jaw, and ribs were all broken?

  • When there are many of things what of the following will be correct [sic]: thing count, thing collection, ... (singular thing), or things count, things collection, ... (plural things) ?

  • Examples: Shop at online shopping stores. Shop on online shopping stores. Given the examples, I hope you get the idea of what I am trying to say. How do I rephrase the sentence to better reflect what I am trying to say—while still keeping it simple and short.

  • what is the difference between “ forever and always” and “ always and forever”?

  • What is the most common or correct spelling of "blah blah blah"?2

And, supposedly, the 3-of-15 statistic implies that the problem is with people who are too hasty to downvote? What about the problem of people who are too flippant about asking questions?

By the way, I've cast over 1,200 votes here, and only about 250 have been downvotes3, so this isn't coming from a serial downvoter. But I thought many of yesterday's questions were well-deserving of downvotes. If we instead welcome such drivel, such drivel will become the norm.


1I understand that not everyone here is a native speaker, so I'm willing to cut some slack some occasional grammatical problems; however, in this case, all these problems had a synergistic effect, to the point where the question as a whole was just downright pathetic.

2Amazingly, this got 9 upvotes, to go with the 9 downvotes.

3Moreover, a good number of my recent downvotes were based on principle, as they were directed toward a suspended user who keeps asking questions despite their suspension.

  • 5
    Thank you J.R. for making the very good point that the OP is responsible for writing a good question, just as the community is responsible for helping cultivate those questions. – Kit Z. Fox Aug 25 '12 at 12:10
  • When I see nothing wrong with my question, what am I supposed to edit or improve? A question of mine was heavily downvoted without even a comment, so I guessedk people are just happy with downvotes and I don't care. – Gigili Aug 26 '12 at 7:08
  • @Gigili: Perhaps you "don't care," but I alluded to your question in a recent meta answer of mine, along with a theory about why it may have been downvoted, in case you're interested. (BTW, I did not downvote your question, so what I wrote there is only a theory.) One other slight clarification: I'm never "happy" with downvotes; I'm much more happy when I can use the upvote button instead. I think most ELU regulars would feel the same way about that. – J.R. Aug 26 '12 at 11:18
  • JFYI, I Googled and even asked a native English speaker about it. Then I decided to ask on ELU. I don't know what your definition of showing efforts or research is. – Gigili Aug 26 '12 at 11:31
  • 1
    @Gigili: Why not edit the question, tell everyone what you Googled, and also add what the native speaker told you? Such additions don't make a question bulletproof to downvotes, but they do go a long way. – J.R. Aug 26 '12 at 11:46
  • 1
    OK, I can do that. Thanks for your input. – Gigili Aug 26 '12 at 11:50
  • @J.R.: I agree with the spirit of your answer in that I think the site should not be filled with bad questions (or answers). But I don't see how downvoting helps this. If someone doesn't have time to edit a post to improve it, is it possible to flag it for editing by others who are willing? I don't know much about the flagging system; I presume this is not currently an option, but maybe it would be useful? (I'll go educate myself on this a bit.) – Rachel Aug 26 '12 at 22:11
  • @Rachel: Flagging asks a moderator to look at an issue. I think it's more severe than a downvote, or vote to close. If I don't like something, I can downvote; if I really don't like it, and I'm sure the community would find it offensive, then I'll flag it for moderator attention. Remember, there are plenty of other websites, like Answers.com, where people can ask anything, and nothing gets downvoted. Downvoting is designed to make people more careful about posting – shoddy posts deserve quick downvotes. I'm not saying there aren't abuses, but I don't think it's as bad as people think. – J.R. Aug 26 '12 at 23:04
  • @J.R.: Right, I know it's used to alert moderators, but I think something like flagging would be good to alert other users that a post needs and deserves the kind of improvement that everyone or almost everyone can make, e.g., editing. So everyone can just check the "deserves improvement" list when they feel like improving posts. I suppose there are rep requirements for editing rather than just suggesting an edit, but even in this case, the work can be shared among more people (assuming it's usually less work to approve an edit than to suggest one). – Rachel Aug 27 '12 at 0:05
  • 1
    @Rachel: A suggested edit will trigger an alert for all users who have enough rep to approve an edit. So long as the edit is a well-crafted improvement, it only takes a matter of seconds to click Accept, and I think most approvers are grateful for the work that has been done (I know I am). I'd like to reemphasize, though, that too much reluctance to use the downvote button could result in a drop in quality, whereby a handful of proactive users are stuck making edits for those who are no longer deterred from posting schlock. – J.R. Aug 27 '12 at 0:35
  • Although I agree that bad questions need downvoting some of your comments seem strange to me. 2. Worried about a space before a question mark?!*% 4. Everybody knows you can't look up preposition usage in a dictionary. 6. This is a standard sentence of the type found in logic and semantics books. 7. a) What's it to you? b) Are any of the questions on ELU important? – Araucaria Jul 23 '14 at 12:08
  • @Araucaria - Just to be clear: #2) I noticed the spacing, but wasn't too worried about it. #4) I agree it's hard to get all the nuances of a preposition from a dictionary; my point was that the O.P. doesn't mention any kind of research – not even an effort . #6) The sentence seems more confusing than "standard". #7) My point is that questions are often better received when they are put into some larger context – it could be important, it wouldn't hurt to explain why. All told, I think this could be improved, and I wouldn't upvote it. Nowadays, I'd suggest it be asked on English Language Learners. – J.R. Jul 23 '14 at 15:50
  • Anyhow, all I meant to say was, before we lament a large number of downvoted questions, we ought to at least examine the questions. I picked one at random, and didn't find the downvotes to be a gross injustice. I realize that some of my points are very nitpicky. (Incidentally, I didn't downvote the question, I only analyzed it.) – J.R. Jul 23 '14 at 15:51
6

I know most people don't agree - but even though I personally don't do it very often, I see nothing wrong with downvoting questions you don't think are likely to be an asset to the site in future. And at least if you downvote, you can reverse later if the question gets improved and/or you reconsider. Once you vote to close, you can't retract.

My gut feel is that over the 18 months I've been here on ELU, there's been a steady increase in the percentage of very basic "English as second language" questions. But I know it's not always easy to recognise complex issues underpinning superficially trivial questions - I learn something from What is an adverb?, but it almost got closed as General Reference within minutes of being asked.

I'd like to see English Language Learners and Linguistics move forward, and I'd like to see ELU act more collectively and coherently to stake out the middle ground between those two.

  • You can "retract" a close vote by voting to reopen if the question gets closed. – Kit Z. Fox Aug 25 '12 at 12:01
6

A Survey of Means, Meaning, and Meanness

Summary:

Here are all SE sites, each with data from the 50 newest questions, and sorted on mean then median. Read-time data was gathered around 01:00 UTC on Sunday, August 26th, 2012. Here’s the tl;dr for those who don’t care to peer more deeply:

  • ELU has a fairly low mean at rank #83/90, but if you notice who comes before us, this isn’t saying as much as one might think.
  • ELU is tied for second place with Academia.SE for having the highest median.
  • ELU is ranked #18/90 in its standard deviation.
  • ELU has a unique mode, with scores −1 and −2 being tied for the most votes at 8 apiece. I do not know how to sort on mode.
Site                              Min  Max    Mean  Median  Stdev  Mode
===============================================================================
SharePoint                         -4    1   -0.02   0.0    0.62   46 × 0
WordPress                          -1    2    0.00   1.0    0.45   43 × 0
Server Fault                       -2    1    0.04   0.5    0.49   41 × 0
Stack Overflow                     -2    2    0.08   0.5    0.59   41 × 0
Game Development                   -8    9    0.14   0.0    2.33   17 × 0
Drupal Answers                     -1    2    0.14   0.5    0.49   43 × 0
Web Applications                   -4    2    0.16   0.5    1.30   23 × 0
English Language and Usage         -5   14    0.18   8.0    3.34    8 × (-2, -1)
Super User                         -2    3    0.20   0.0    0.72   36 × 0
Meta Stack Overflow               -14   19    0.30  -5.0    5.23   12 × 0
Ask Different                      -2    6    0.44   0.5    1.15   29 × 0
SmugMug                            -1    2    0.46   1.0    0.75   26 × 0
Ask Ubuntu                         -1    6    0.52   1.5    1.20   30 × 0
OnStartups                         -3    9    0.58  -1.5    1.78   20 × 0
Webmasters                         -2    5    0.74   0.0    1.61   30 × 0
Quantitative Finance               -5   11    0.88   7.0    3.69    7 × (3, 0)
Android Enthusiasts                -4    5    0.90   0.5    1.76   22 × 0
Cryptography                       -4    5    0.90  -2.5    1.72   16 × 0
Software Quality Assurance and Te   0    4    0.92   0.0    1.09   22 × 0
Audio-Video Production             -1    5    0.98   1.5    1.17   21 × 0
Electrical Engineering             -4    7    1.00   2.5    2.30   16 × 0
Arqade                             -6    6    1.02   2.0    2.01   15 × 1
Database Administrators            -2    4    1.06   0.5    1.16   17 × 1
Programmers                        -5   11    1.18   0.5    3.69   10 × 0
Graphic Design                     -2    6    1.20   1.0    1.61   19 × 0
Personal Finance and Money         -1    6    1.24   0.5    1.42   16 × 0
Code Review                        -1    5    1.26   2.0    1.48   19 × 0
Signal Processing                  -2    6    1.26   2.5    1.94   19 × 0
Physics                            -4    8    1.34   2.5    1.91   16 × 1
GIS                                -3    8    1.36   1.0    2.20   15 × 0
Mathematics                         0    7    1.46   2.0    1.77   19 × 0
Cross Validated                     0    5    1.52   1.5    1.42   16 × 0
Home Improvement                   -1    6    1.56   1.0    1.56   15 × 0
Unix and Linux                     -2    6    1.62   1.0    1.41   15 × 1
Philosophy                         -3    7    1.64   0.5    2.23   14 × 0
Salesforce                         -7    7    1.64   5.0    2.23   14 × 1
IT Security                        -5    8    1.68  -2.5    2.22   12 × 0
Sports                             -6   10    1.74  -5.0    3.07   11 × 3
Chemistry                          -3    7    1.80   1.5    2.09   12 × 1
Motor Vehicle Maintenance and Rep   0    6    1.92   3.5    1.25   14 × (1, 2)
Bicycles                           -4   13    1.98  -2.0    2.86   10 × 1
Writers                            -3    9    2.00   2.0    2.28   10 × 0
Physical Fitness                   -2    8    2.10   2.0    1.75   12 × (3, 2)
Stack Apps                          0   12    2.20   0.5    2.49   14 × 1
Homebrewing                         0    9    2.24   2.5    1.80   11 × 2
Bitcoin                             0    8    2.24   6.0    1.59   16 × 3
User Experience                    -3   25    2.28   3.0    4.19   12 × 1
Poker                              -2    9    2.32   2.5    2.01   11 × 3
TeX - LaTeX                         0    7    2.34   4.5    1.89   14 × 1
Chinese Language and Usage         -1    9    2.40   2.0    2.08   12 × (2, 1)
Movies and TV                      -6    9    2.40  -2.0    2.62   12 × 3
Gardening and Landscaping          -2   14    2.42   4.0    2.40   13 × 2
Mi Yodeya                          -6   10    2.58   1.0    2.93    9 × (2, 0)
Computer Science                   -4   11    2.60   3.0    2.52   13 × 2
Photography                        -1   12    2.62   2.5    2.81   10 × 3
Windows Phone                      -1    7    2.68   4.5    1.70   15 × 3
History                            -6   13    2.70   1.0    3.60    9 × 2
Board and Card Games               -1    8    2.70   3.0    2.19   11 × 1
Seasoned Advice                    -4   18    2.82   0.5    3.42    9 × 3
Personal Productivity              -1    8    2.82   3.0    2.52   16 × 1
Russian Language and Usage          0    7    2.84   3.5    1.65   14 × 2
Islam                              -2    9    2.90   3.0    2.05   16 × 3
Musical Practice and Performance   -3   10    2.96   3.5    3.03    8 × (3, 5)
Theoretical Computer Science       -5   20    3.24   7.0    5.58    7 × 0
Biblical Hermeneutics              -1    8    3.32   3.0    2.02   12 × 4
Parenting                          -1   12    3.34   2.0    2.75   12 × 3
Biology                            -1   12    3.42   4.5    2.78    9 × 4
Project Management                 -4   10    3.46   3.5    2.81    8 × (3, 1)
Computational Science              -1    9    3.46   5.5    2.33    8 × (3, 2)
Linguistics                         0   22    3.52   4.0    3.63   11 × (3, 2)
Martial Arts                       -1    8    3.60   0.5    2.08   11 × (3, 4)
Spanish Language and Usage         -2   16    3.76   4.5    2.81   16 × 3
The Great Outdoors                 -3   10    3.78   4.5    2.21   10 × 5
Christianity                       -4   12    3.82   4.5    3.04   13 × 4
Raspberry Pi                        1   10    3.98   3.0    2.40   11 × 2
Cognitive Sciences                 -4   26    4.04   2.0    4.46    7 × (6, 3)
Mathematica                        -9   17    4.06   4.0    5.08    9 × 2
Code Golf                          -3   22    4.22   0.0    5.13    7 × 0
German Language and Usage          -2   19    4.32   1.0    3.78   13 × 2
Role-playing Games                 -6   26    4.38   2.5    4.37   11 × 3
Libraries and Information Science   0   12    4.42   7.5    2.64   10 × 3
Chess                              -1   12    4.46   4.0    2.47   12 × 4
Travel                              0   21    4.50   4.5    3.24    9 × 4
French Language and Usage           0   10    4.50   5.0    2.44   10 × 6
Japanese Language and Usage         0   13    4.92   2.5    2.93    8 × 6
LEGO® Answers                       1   11    4.96   4.5    2.79   11 × 3
Science Fiction and Fantasy         0   18    5.16   0.5    4.04    9 × 3
Skeptics                           -4   90    5.88   7.5   12.78    7 × (-1, 5)
Academia                           -3   32    6.40   8.0    5.60    9 × 5
The Workplace                      -7   36    8.42   8.5    7.74    6 × (7, 12, 5)

Shell script available upon email request; requires perl because I’m me, and wget because I’m lazy. Site list is dynamically generates as new SE sites added.


Old Posting

I don’t believe that ELU is especially different in this compared to other StackExchange sites of its type. In particular, both the mean and median score of questions on the ELU front page are each comparatively high, but so too is our standard deviation and the ratio of our standard deviation to our mean.

We also have a curious mode: only one other site besides ELU has a mode (most frequent score) that includes a negative score, and that is the Skeptics.SE site. We have 6 questions at −2, while they had 4 at −1. It makes you wonder what sort of questions they draw, and why they are like us in this regard. However, Skeptics.SE also has the highest mean of all sites surveyed, so they are different from us in that regard.

Meanwhile, here is the same dataset sorted a variety of ways. The primary sort column is starred.


Sorted on Mean

Here are a bunch of SE sites’ stats as of a snapshot from a little while ago, sorted on ascending mean:

Site           N  Min  Max   *Mean Median   Stdev   Std/Mean Mode
======================================================================
so            90   -3    6    0.46   1.00    1.30    2.86    58 × 0
wordpress     48    0    4    0.50   0.00    0.94    1.87    34 × 0
superuser     31   -1    9    0.55   0.00    1.64    3.00    21 × 0
ask-ubuntu    48    0    7    0.83   0.50    1.33    1.59    27 × 0
webmasters    48   -2    5    0.85   0.00    1.46    1.71    25 × 0
apple         48    0   12    1.35   1.00    2.21    1.63    21 × 0
dba           48   -2   10    1.54   1.50    2.19    1.42    16 × 0
physics       48    0   20    2.40   2.50    3.11    1.30    13 × 1
serverfault   48   -1   76    2.62   0.50   11.09    4.23    25 × 0
it-sec        47    0   21    3.38   7.00    3.82    1.13     9 × 3
math          48    0   16    3.73   8.50    4.19    1.12    13 × 1
arcade        48   -3   23    3.73   3.00    4.77    1.28    12 × 1
elu           46   -3   37    4.04   3.00    7.02    1.73     6 × (-2, 0, 2)
elu-meta      43   -6   29    4.37  -3.00    6.78    1.55     7 × 1
spanish       48   -2   20    4.48   2.50    3.81    0.85    12 × 3
german        48   -2   19    5.21   8.00    4.48    0.86    11 × 2
tex           48    0   28    5.35   6.00    6.20    1.16     9 × 4
cs-theory     44   -3   25    5.55   4.50    5.96    1.08     5 × (0, 4)
scifi         38    0   30    6.26   7.50    5.54    0.88     5 × (3, 4, 5)
codegolf      48   -3   29    7.77   3.00    7.77    1.00     6 × 4
rpg           48   -1   46    8.48   3.00    9.00    1.06     8 × 3
programmers   34    0   65   10.38  17.50   14.52    1.40     5 × 6
so-meta       46   -5  175   11.46   1.50   32.65    2.85    13 × 0
skeptics      46   -1   90   12.91  15.50   15.38    1.19     4 × (-1, 5)
======================================================================
Site           N  Min  Max   *Mean Median   Stdev   Std/Mean Mode

Sorted on Median

Even if you sort on the median score, defined as half being above and half being below, then again, ELU comes out much better than many more frequently visited sites:

Site           N  Min  Max   Mean *Median   Stdev   Std/Mean Mode
======================================================================
wordpress     48    0    4    0.50   0.00    0.94    1.87    34 × 0
superuser     31   -1    9    0.55   0.00    1.64    3.00    21 × 0
webmasters    48   -2    5    0.85   0.00    1.46    1.71    25 × 0
ask-ubuntu    48    0    7    0.83   0.50    1.33    1.59    27 × 0
serverfault   48   -1   76    2.62   0.50   11.09    4.23    25 × 0
so            90   -3    6    0.46   1.00    1.30    2.86    58 × 0
apple         48    0   12    1.35   1.00    2.21    1.63    21 × 0
dba           48   -2   10    1.54   1.50    2.19    1.42    16 × 0
so-meta       46   -5  175   11.46   1.50   32.65    2.85    13 × 0
physics       48    0   20    2.40   2.50    3.11    1.30    13 × 1
spanish       48   -2   20    4.48   2.50    3.81    0.85    12 × 3
arcade        48   -3   23    3.73   3.00    4.77    1.28    12 × 1
elu           46   -3   37    4.04   3.00    7.02    1.73     6 × (-2, 0, 2)
codegolf      48   -3   29    7.77   3.00    7.77    1.00     6 × 4
rpg           48   -1   46    8.48   3.00    9.00    1.06     8 × 3
cs-theory     44   -3   25    5.55   4.50    5.96    1.08     5 × (0, 4)
tex           48    0   28    5.35   6.00    6.20    1.16     9 × 4
it-sec        47    0   21    3.38   7.00    3.82    1.13     9 × 3
scifi         38    0   30    6.26   7.50    5.54    0.88     5 × (3, 4, 5)
german        48   -2   19    5.21   8.00    4.48    0.86    11 × 2
math          48    0   16    3.73   8.50    4.19    1.12    13 × 1
skeptics      46   -1   90   12.91  15.50   15.38    1.19     4 × (-1, 5)
programmers   34    0   65   10.38  17.50   14.52    1.40     5 × 6
======================================================================
Site           N  Min  Max   Mean *Median   Stdev   Std/Mean Mode

Sorted on Standard Deviation

One area were we come up high is our standard deviation compared with that of other sites, and not unsuprisingly, also the ratio of the standard deviation to the mean. Here is the same dataset sorted on ascending standard deviation:

Site           N  Min  Max    Mean Median  *Stdev   Std/Mean Mode
======================================================================
wordpress     48    0    4    0.50   0.00    0.94    1.87    34 × 0
so            90   -3    6    0.46   1.00    1.30    2.86    58 × 0
ask-ubuntu    48    0    7    0.83   0.50    1.33    1.59    27 × 0
webmasters    48   -2    5    0.85   0.00    1.46    1.71    25 × 0
superuser     31   -1    9    0.55   0.00    1.64    3.00    21 × 0
dba           48   -2   10    1.54   1.50    2.19    1.42    16 × 0
apple         48    0   12    1.35   1.00    2.21    1.63    21 × 0
physics       48    0   20    2.40   2.50    3.11    1.30    13 × 1
spanish       48   -2   20    4.48   2.50    3.81    0.85    12 × 3
it-sec        47    0   21    3.38   7.00    3.82    1.13     9 × 3
math          48    0   16    3.73   8.50    4.19    1.12    13 × 1
german        48   -2   19    5.21   8.00    4.48    0.86    11 × 2
arcade        48   -3   23    3.73   3.00    4.77    1.28    12 × 1
scifi         38    0   30    6.26   7.50    5.54    0.88     5 × (3, 4, 5)
cs-theory     44   -3   25    5.55   4.50    5.96    1.08     5 × (0, 4)
tex           48    0   28    5.35   6.00    6.20    1.16     9 × 4
elu-meta      43   -6   29    4.37  -3.00    6.78    1.55     7 × 1
elu           46   -3   37    4.04   3.00    7.02    1.73     6 × (-2, 0, 2)
codegolf      48   -3   29    7.77   3.00    7.77    1.00     6 × 4
rpg           48   -1   46    8.48   3.00    9.00    1.06     8 × 3
serverfault   48   -1   76    2.62   0.50   11.09    4.23    25 × 0
programmers   34    0   65   10.38  17.50   14.52    1.40     5 × 6
skeptics      46   -1   90   12.91  15.50   15.38    1.19     4 × (-1, 5)
so-meta       46   -5  175   11.46   1.50   32.65    2.85    13 × 0
======================================================================
Site           N  Min  Max    Mean Median  *Stdev   Std/Mean Mode

Summary

I don’t see that the data are showing that we are particularly hard on questions — mean, if you would — compared with other sites. I welcome alternate reads of the data.

All data were gathered between 14:00 and 14:30 UTC on Saturday, August 25th, 2012.

  • Are there URLs (vs downloading data and running scripts) by which others can produce similar tabulations? – James Waldby - jwpat7 Aug 25 '12 at 20:39
  • @jwpat7 You can try the StackExchange Data Explorer. I'm not sure how much data you have access to that way though. – Kit Z. Fox Aug 25 '12 at 21:30
  • @KitFox - Yes, I've used data.stackexchange for a couple dozen queries in the past year, but I think it doesn't support queries of the form in tchrist's tables. I'm supposing a Creative Commons Data Dump plus some DB or stats program is needed for the tables. Note, I regard the displayed stats as irrelevant to OP, and not interesting or illuminating. (They do appear to support the various points tchrist listed.) The question I have for tchrist is how the data was fetched and processed. – James Waldby - jwpat7 Aug 26 '12 at 6:25
  • @jwpat7 headslap I deleted a comment of tchrist's (and one of mine) but of course, this relevant part was for you, now that I think about it: "If you send me mail, I’ll send you the scripts to run it yourself, since this changes all the time. But not much." – Kit Z. Fox Aug 26 '12 at 11:22
  • @tchrist This is just awesome. Have the meta types on meta SO seen this? Don't concern yourself regarding the lack of sort on the median. You have done more than enough already. This answer is enlightening but not revelatory (a revelation? I see pesky redline...) for EL&U, but your tables should provide the meta SO crowd, or even Area 51, plenty of material from which to draw insights. – Ellie Kesselman Aug 27 '12 at 6:27
6

Downvotes by themselves are not a problem, but they might point you to a larger issue. I can speculate as to some potential causes, but I don't know how to validate them:

1) Was there a corresponding uptick in bad questions?

2) Is there a serial-downvoter just out there causing havoc?

3) Is the recent "Community Review" system bringing more attention to bad questions than good ones on the "recently active" page?

4) Are the users of ELU just reaching a breaking point of frustration with the influx of silly, trivial questions and downvoting everything even marginally off-target?

I think it may be a little of each, but it's the latter that I find most concerning, and is probably the most difficult to fix.

  • 5
    +1 for #4. I think you may be onto something there. – J.R. Aug 27 '12 at 0:37

You must log in to answer this question.

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