10

I have been traveling the past few weeks so I haven't kept up with the election stuff. I did look at the candidates and they all seem like they would do a good job to me. Who knows one might be corrupt and turn this into a Sanskrit shop?

But I feel that people have been beating around the bush a bit on the issue with the site. I wrote an answer-question here which I will ask here since it will allow people to answer more appropriately.

This is a bit of an extension of John/FF's questions... If ELU stays as a separate entity as ELL, how do we ensure that the people answering the questions don't belong on the ELL?

It is pretty easy to moderate questions and close them off. But I find that the biggest issue with the site is the users answering. Being a long-time SE user and one that uses many SE sites I get that you will have people answering questions without the "right background".

What happens is you get really bad answers. Those answers get downvoted and the user flails. But that doesn't work here because of the ELL/ELU issue (as pointed out by FF's question).

We have several users that are obviously not native speakers (John points this out) that answer a high volume of questions. Some of these users have been near the top of the leaderboard lately. They answer everything, seemingly using reverse dictionaries. They edit their answers 20 times based on comments and after finding out the first 10 things they put were incorrect. Often the 20th thing isn't much better or isn't explained with the right context.

Often they take answers from other users or simply take another answer and use a thesaurus to answer a question. This isn't rocket science. We all know who these users are.

So my question is what do the candidates think about these users? What can we do about this situation?

Personally I come to the site to learn. I find some of the accepted answers and highly upvoted answers ridiculous at times - so much so that I am not learning, I am doubting my own knowledge. I feel that these "answers" detract from the site and also halt good answers/discussion. These answers usually come with some sort of "proof" that taken by itself might answer the question but put in context they are absolutely incorrect.

Not sure how other people feel - but I just get tired of reading a half page of "research" from a non-native speaker doing a reverse dictionary on stuff that just doesn't make sense. I can downvote these things and I do. And I can upvote the good answers buried underneath and I do. But that doesn't make up for the 20 nonnative speakers that already upvoted the wrong answer, which goes back to FF's question.

And this kind of gets into some of the topics phenry discussed in his recent meta question. The bad questions can be tackled with different bylaws and policing but there is nothing in SE to combat users who shouldn't be answering and users who shouldn't be voting. Well I know there are some things in place but they don't work here.

14
  • You should also visit the election chat room and ask there too. I mentioned your name yesterday, you must have heard me!
    – Mari-Lou A
    Jun 20, 2014 at 21:12
  • 1
    Just a quick idea: how about needing to earn 500 - 1000 rep from answers on ELL before you are qualified to answer questions on ELU? 500 rep would require 50 upvotes, which should be enough to prove that a user knows what they're talking about.
    – Dom
    Jun 21, 2014 at 17:14
  • 5
    @user that would be awful, it would take much longer to earn 1000 rep on ELL than it does on ELU, and the type of questions asked are often mind blowing dull or obtuse. Not all, but many. I've practically withdrawn from that site, again, it's not easy to earn rep points on ELL, the questions are very basic = boring, and you often have to be a grammar queen or king to answer those queries.
    – Mari-Lou A
    Jun 22, 2014 at 9:27
  • @Mari-LouA point taken, just an idea. I don't know much about this issue, I've just seen it popping up. Haven't been past the homepage on ELL and I only really do single word/phrase requests here. I agree that there is an issue of high votes scores on low quality answers though, I raised an issue about it myself on GD SE, and read another similar post on parenting SE meta.
    – Dom
    Jun 22, 2014 at 9:43
  • 1
    @user if you also noticed the same thing occurring in different SE sites, then it's no longer a matter of non-native speakers not knowing any better but it being more connected with questions hitting the hot page, gaining notoriety, and an answer being mass voted. That happens frequently on EL&U too.
    – Mari-Lou A
    Jun 22, 2014 at 10:11
  • @Ryebread - I added a fifth option to my answer. I'm slightly surprised that your question has only drawn a single response from a moderator-candidate and none at all from an existing moderator. Maybe just weekendiness ?
    – Frank
    Jun 22, 2014 at 10:56
  • @Mari-LouA - I think some of the issue is the Hot Page. However like you said this affects all sites and they all get non-expert voters. I think the biggest issue is with the non-native (+non-expert because I know some non-natives do have the expertise to answer here) answerers. It just makes the whole site silly. It's like why would I bother answering something that has a ridiculous answer with 10 upvotes. I don't care about points but I do care about wasting my time. Jun 22, 2014 at 23:12
  • I agree the quality on this site is lacking and it makes me lose interest in contributing. Maybe it could be possible to require all answers to provide some sort of evidence for the usages they propose? Jun 23, 2014 at 2:15
  • @Frank I have urged Ryebread to visit at least once the election chat room, there he will find; Andrew Leach, Medica, Phenry, Matt Эллен and Mr Hen. If you ping his name, Yoichi Oishi will also reply to questions. The other candidates are virtually absent. If he cares so much about this topic then now is the time to get answers from the site's THREE mods, unless he decides to vote for a candidate who appears to be indifferent to the community's questions. (That is my provocation aimed more at the "absentee" candidates than to Ryebread)
    – Mari-Lou A
    Jun 23, 2014 at 4:18
  • @RyeɃreḁd - I'm willing to distill this into a question and post it into the ELU election Blog. But, since that is precisely where you're supposed to post questions, not necessarily here in meta during an election, I won't answer this here. You'll get pinged with answers. If I missed anything, please follow up in Election Chat. Jun 23, 2014 at 6:00
  • @RyeɃreḁd - I have posed your questions and they are being answered in Election Chat already. Come read the answers, and pose follow-ups if you want. Jun 23, 2014 at 8:22
  • Btw, your earlier question was discussed here on chat: chat.stackexchange.com/transcript/14995?m=16131820#16131820
    – MrHen
    Jun 23, 2014 at 13:16
  • @MrHen - thanks for the link. Sorry for the delay but I was in Mari-Lou's neck of the world. Good to see the concern but don't think we have an answer for this yet. Jun 27, 2014 at 19:24
  • @medica - thank you for posting. Wish we would get heavy users or mods answering this other than Mari-Lou who did a great job with her answer. I provided a vision of what I would do (and I could actually do this on the SE clone I have at my company). As it stands I don't know if my answer is doable or within SE vision. [at my company people just get called a dumbass when trying to farm for votes so never had to think about it, yet] Jun 27, 2014 at 19:27

3 Answers 3

10

We have several users that are obviously not native speakers (John points this out) that answer a high volume of questions. Some of these users have been near the top of the leaderboard lately. They answer everything, seemingly using reverse dictionaries. They edit their answers 20 times based on comments and after finding out the first 10 things they put were incorrect. Often the 20th thing isn't much better or isn't explained with the right context.

I know exactly who you are talking about, but it's one/two non native speakers out of hundreds/thousands, one particular user does edit his/her answer twenty (and once, to my amusement, even thirty-three times!) but at least he/she was aware that the post needed improvement. And who's to say that the twentieth edit is not an improvement? On the contrary, I have found it has been. Does it mean I'll upvote it, not necessarily. Besides, what about the times when the users posts are pretty good? When the answers suggested do fit, what do you do then? Do you refuse to upvote because of their previous track record?

I remember the OP's first contributions weren't exactly pearls of wisdom, however the quality of his posts over time has improved dramatically and I enjoy reading his contributions, he is someone who I can rely on to get an American perspective. I trust his judgement (read answers), but I needn't always agree.

What EL&U needs (and relatively soon) are more native speakers who ONLY vote, they needn't be grammar gods, but they would instantly see if an answer or question is grammatical without necessarily understanding why. The problem arises though when those "English native speakers" attempt to answer grammar-type questions. Often they'll just provide the "answer" without any explanation, and completely disregard any alternative solutions. They'll fail to explain "why", and the number of times I've seen misspellings, poor capitalization, and poorly constructed sentences from native speakers is embarrassing. Being a native speaker does not mean you have an awareness of your language, how it functions, how the words connect with each other; whereas speakers of English as a second (or third) language often do, and will want to understand further by asking those type of questions which your average native speaker will never have reflected upon. Questions which can be stimulating and fun to answer but will mistakenly be dismissed as being trivial, unimportant and general reference because your average NS will only be able to reply: "because it is".

An English language site populated and governed by the greater part by English native speakers is one destined to die. Don't believe me, look at SE Linguistics, practically a graveyard judging by the number of users and questions asked. If the OP wants a site where his intellectual prowess can be tested, I'd recommend that he visit that site. He might find it stimulating and gratifying or... out of his depth.


EDIT
My suggestion that EL&U needs native speaker voters, is a provocation, nevertheless it is true that there is a rise in the number of inappropriate answers being upvoted. If the more seasoned members of this community wish to counteract this phenomenon there are FIVE available recourses:

1) Upvote the best answers.

2) Downvote those answers which are blatantly incorrect.

3) Leave a comment at the post (not necessarily accompanied by a downvote) explaining where the poster is mistaken, or how the post has not answered the OP's question.
This is the approach I prefer, and often posters have made edits which IMO have improved their posts. Consequently, I delete my comments and, sometimes, upvote their contributions. I do NOT provide an "answer". If the poster disagrees, and I really disagree with his/her answer I leave my comment(s).

4) Edit answers: a) correcting obvious grammar mistakes b) improve formating (paragraphs, italics in place of caps etc.) to improve overall legibility.

5) Write a better answer.
This I have done myself when the OP (in this case always a non native speaker) has accepted an inappropriate answer five minutes after it has been posted. Sometimes I leave a comment suggesting the OP waits at least a couple of hours before accepting the first posted answer. If my own post is upvoted once, twice etc. and other users have contributed, the OP can make a more informed decision.

8
  • @Jack Who and how would someone go about contacting the user(s) in question? PMs (personal messages are not available to users, not everyone wants to or feels the need to visit chat, and to my limited understanding, chat rooms are visible to all—so no chance of privacy. Meta is an appropriate place, as long as user names aren't mentioned. And I wouldn't class Ryebread's question as being "gossip". He's talking about posts, and how the system can be abused.
    – Mari-Lou A
    Jun 25, 2014 at 10:23
  • Except for the second one, I think these are all good ideas, and I do them all. Except I don't like to downvote, because it's uninformative, and it's normally uncalled-for except in cases of simple vindictiveness. I think one feature that could be tested is having one's native language(s) required in signing up, and appearing on their home page. If people felt it was a good idea, professional qualifications might also be optional additions. In any event, an answer by a native speaker should be understood differently from one by a non-native; and the same is true of questioners. Jun 25, 2014 at 19:54
  • @JohnLawler I believe that would take a bit more programming from the SE side of things (as far as I know that isn't an existing feature of the API), and would likely be declined if brought up as a feature request on Meta. The solution for native/non-native identification is to encourage users to add that information to their profile (imo). However, there are some that would see that as a form of segregation, e.g. I have to identify myself as a native French speaker on English.SE? Do they not trust my posts? I could see that becoming a reason for some users to leave the site... Jun 26, 2014 at 21:23
  • 1
    I don't trust anybody's posts. French speakers are not really a problem. Asian Englishes, on the other hand, are much more divergent from native dialects. They're all English, but there's a whole world out there, and questions that come up naturally in one context make no sense at all in another. Giving context is something we should encourage. Just like giving several examples of a phenomenon instead of trying to describe it in broken grammar-speak. Jun 26, 2014 at 21:39
  • 1
    @JohnLawler - I don't think it is even their dialect. And it really isn't even that a non-native answering a question. As Mari-Lou states sometimes they are "correct". It is the bomb of answers we get from these people. Obviously the downvoting doesn't work because we still get crap answers with 10 upvotes while a good one has 2. And for the most part it is a couple users but there is a new non-native speaker every week or two that bombs the site with answers... I don't know. I really hate complaining without a solution. [ Jun 27, 2014 at 18:21
  • I accepted yours because I thoroughly agree with everything you bring up. Although there is no answer for the voter with poor English skills. I did provide an alternative answer but I think there would have to be slight tweaking to SE backend so not sure it is feasible. Jun 27, 2014 at 18:32
  • @RyeɃreḁd we still get crap answers with 10 upvotes while a good one has 2 I doubt you'll ever prevent people up voting a crap answer, but why would a good answer only have 2 up votes? Is it that those who can see what the good answer is just aren't voting them up? Maybe you need to turn your attention to the stalwarts not the transients.
    – Frank
    Jun 28, 2014 at 15:27
  • 2
    @Frank - went through a big set of questions yesterday to read/review. There were 4 distinct users that had 5K reputation with ridiculous answers. Now some people were calling them out and I deleted a few... but SE should figure this out so there isn't hostility in the comments. Jun 28, 2014 at 16:37
3

So my question is what do the candidates think about these users? What can we do about this situation?

The basic SE model applies to all SE sites

Anybody can ask a question

Anybody can answer

The best answers are voted up and rise to the top

Regarding voting (up or down) on an answer

...good content rises to the top

...incorrect content falls to the bottom

On accepted answers

Accepting an answer is not meant to be a definitive and final statement indicating that the question has now been answered perfectly. It simply means that the author received an answer that worked for him or her personally...

Assuming that you want to stay within the general SE framework

  1. You could suggest a change to the reputation limits perhaps raise the bar on answering a question to, perhaps, 5000. That would prevent the 'unproven' users from providing answers but would also alienate new users who could provide answers that meet your quality standards. At that level the 'answerer' pool would be reduced to around 100 users.

  2. You could suggest that all users must declare what sort of education they have and/or whether they class themselves as native or non-native speakers and have a moderator judge them accordingly whether they are allowed to answer or not. That doesn't sound like it would be an acceptable SE policy but you never know but I suspect it would be difficult to manage as people would presumably need to provide evidence for their claims.

  3. Moderators could be given the freedom to simply delete answers they judge as wrong, then those answers couldn't get votes.

  4. You could down vote answers that you believe are wrong, leave a comment explaining why it's wrong and provide what you believe to be the correct answer. If the question author still selects a 'wrong' answer as the accepted answer then that is their choice. Perhaps even more important is to up vote all answers you feel are right/good.

  5. Users could be prevented from both asking and answering questions until they have read the What is General Reference page. That functionality is available in SE through badges I believe. The GR page is something of a shambles but it could easily be tidied up into a proper If your question can be answered through any one of these linked sites then it will be closed as off-topic. If you answer comes from any one of these linked sites then the answer is a poor answer as the question author could have found it themselves, do not post answers that come from General Reference sites. Perhaps editing it into a list and locking it and only allowing moderators to add new items to the list would help maintain a structure that people might actually look at before asking or answering.

Items 1 & 2 will almost certainly draw more accusations of elitism/snobbery and whatever else EL&U is renowned for but isn't that the point? If I asked a question here I'd want it answered by someone who at least claims to know what they are talking about, not someone who happens to have a dictionary to hand.

Item 3 may have the [unintended] consequence of branding EL&U as unfriendly.

If item 4 doesn't 'work' then that might suggest that EL&U is not a good fit for the SE platform.

Item 5 will reduce the activity on the site massively, possibly to the point of making it a useless SE site

As an aside

The term 'non-native speakers' get used a lot on this site but it seems to be used to mean 'not very good [by some intangible standard] at English'. ELL makes it very clear that it welcomes non-native speakers. EL&U does not make it clear that it wants (if indeed it does) to discourage non-native speakers and/or people without a very good grasp of English; perhaps it should.

I'd also like to ask how you know (apart from 'I just know') that the answers you are talking about have been written by non-native speakers and have been up voted by non-native speakers ?

In summary

What do the candidates think about these users?

I'm not a candidate and I don't know who 'these users' are.

What can we do about this situation?

Item 4 in the list above would be my choice.

10
  • 2
    I concur. We need to work within the SE model. And that means voting to weed out poor questions (mainly up and down, but also to put poor ones on hold pending clarification/further detail) and voting on answers so good ones rise and poor ones don't. Did I mention voting? More on this in Chat at this message.
    – Andrew Leach Mod
    Jun 21, 2014 at 12:10
  • @AndrewLeach - I don't want to speak in absolutes but the ELU will probably always be a very basic English site if there no rules beyond SE norms. Then might as well merge ELL and ELU. There are obviously TONS of ELL users coming to the site to post questions, and those same users are voting on answers (some answering). Frank brings up some good ideas. I don't know that there is a perfect idea or I would have given my thoughts. The basic voting isn't working because there seems to be too high a volume of non-native English speakers voting. Jun 22, 2014 at 3:27
  • I like the first part of your answer, a summary on how the SE model works. But, as for this proposal: You could suggest a change to the reputation limits perhaps raise the bar on answering a question to, perhaps, 5000. How in the world is anyone going to reach 5000 rep points if they are not allowed to answer questions?
    – J.R.
    Jun 27, 2014 at 9:01
  • As for this: EL&U does not make it clear that it wants (if indeed it does) to discourage non-native speakers and/or people without a very good grasp of English. On the contrary, ELU makes it very clear what it wants: linguists, etymologists, and serious English language enthusiasts. Non-native speakers are welcome of course, but anyone (native or not) who consistently doles out inaccurate information and shoddy advice will wear out that welcome in due time.
    – J.R.
    Jun 27, 2014 at 9:07
  • @JR The 5000 bar, applied now, would leave EL&U with at least 100 users allowed to answer; part of the problem with doing that, noted below it in italics, is that new users will never make the cut regardless of their ability so the site will become stagnant and eventually wither and die.
    – Frank
    Jun 27, 2014 at 9:14
  • 2
    @JR serious English language enthusiasts There's the problem. If I am learning the English language and I'm serious about learning English to the point of being enthusiastic about it, what part of that do I not fit into? (unless of course English is meant to mean born and bred in England)
    – Frank
    Jun 27, 2014 at 9:18
  • @Frank - The conjunction is and, not or. (Then again, an English learner isn't so likely to pick up on such a subtle distinction, so that's not likely to solve the problem either.) :^)
    – J.R.
    Jun 30, 2014 at 10:43
  • @J.R. I'm a life long native speaker and I don't read it as linguist and serious... or etymologist and serious .... What hope for the seventy percent of traffic that originates outside US/UK?
    – Frank
    Jun 30, 2014 at 11:14
  • Frank - I was just kidding around. Fact is, there are two audiences out there: one is interested in tackling the more puzzling aspects of English, and the other consists of those who want to be helped as they learn English. Some don't mind the two intermingled; some would rather keep them separate – and I understand why. I like watching my kids play baseball, and I like watching professional baseball, but I'd rather not watch kids play baseball when I'm wanting to watch the pros play. Anyhow, the birth of ELL didn't seem to make the "issue" go away, but the attempt was well-intended.
    – J.R.
    Jun 30, 2014 at 13:10
  • @J.R. Continuing the analogy, the kids play all day every day, the pros only play once a week. Could the site sustain itself on one game a week? I think ELL isn't promoted enough (it is still a Beta) but I wonder if EL&U wouldn't do better as a link off ELL If you are a native English speaker ask your question at EL&U, maybe write it in Old English as a gatekeeper test :). It won't be long until the Indian traffic exceeds the US traffic on EL&U unless ELL becomes the default 'English' StackExchange site.
    – Frank
    Jun 30, 2014 at 13:30
3

Here is my suggestion since I feel the need to have one since I complained.

Have a flag that says something along the lines "This answer reflects a lack of basic understanding of the English language". Let users over a certain criteria vote if the flag was good/bad.

If users get a certain amount of these flags passed for a given time period the get suspensions in increasing order. Note that editing your question shouldn't impact the flag. Whatever the user saw at the time stands.

So something like this:

10 of these flags in a two month period - 3 days
10 more after - 7 days
10 more after - 14 days

The point isn't to kick these users off the site, it is so they quit bombing the site with poor answers.

Note: I fully understand that I have given answers that could be flagged for this. I have 5-6 times misread a question and have been called out. I have then promptly deleted answer or dramatically changed answer. Would I be upset about being suspended for this? No. I would probably read questions more thoroughly and not answer when drinking. As it stands, cheers!

4
  • I like the basic idea of such a flag. However, with the format you propose, just one terrible answer that received 10 flags would be enough to prompt a suspension. I think it would be better to base suspensions on something like "the proportion of the member's answers posted in the previous 30 days flagged at least once for demonstrating a lack of basic understanding of the English language". I'd also suggest a cap of (say) 5 votes that any given fellow-member could cast against the alleged offender's postings, in order to make it harder to pursue a systematic vendetta against that person. -->
    – Erik Kowal
    Jun 30, 2014 at 8:44
  • --> This criterion naturally prompts the question "What should that proportion be?" I don't have a ready answer for that; I think it ought to be data-driven. Perhaps it could be arrived at by a panel drawn from the most senior members of the EL&U forum examining the most obvious offenders' existing answers from the past 30 days and subjectively scoring their relative usefulness or uselessness. The statistics derived from that exercise could then be used to set a pass/fail threshold at a level that would have triggered a moratorium on answering for all or most of the offenders scrutinized.
    – Erik Kowal
    Jun 30, 2014 at 8:56
  • @ErikKowal - I was under the assumption that a user's post could only draw one flag (in the end). Jun 30, 2014 at 15:02
  • What I meant was that for the purposes of calculating suspension eligibility, in any given 30-day period there should be a cap on the number of incompetence flags that could be bestowed on one member by another. (The other possible sanctions and countermeasures would still be available, such as downvoting, critical comments or remedial editing.) So in my scheme, in any 30-day period A could flag five postings by B on the grounds of incompetence, but that would be all. B's final score would be dependent on the competence formula and the number of incompetence flags awarded by other members.
    – Erik Kowal
    Jun 30, 2014 at 22:07

You must log in to answer this question.

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