-5

What I mean by the the fundamental difference is that it is very difficult for users on the Japanese language site to accurately confirm the correctness of the responses they are getting.

With sites based on computers or code it is very easy to simply try the solution posted and see if it works. However with language even if you tried it and it seems to work, it doesn't always mean it is correct.

It seems to me that most votes on this site are going out based on a personal feeling of how correct the answer seems to be. Of course the problem with this is that that personal feeling has nothing to do with if that answer is correct or not.

I think it is clearly a big problem if people can not count on the information on this site to be reliable.

So my question is how is this being addressed now and are there any plans to implement any features that may address it in the future?

Also does anyone have any suggestions as to how this can be addressed if there are not currently any plans to do so?

  • 2
    How can you be sure votes are given on a personal feeling only or mostly? – Alenanno Apr 29 '12 at 1:36
  • I am not. However it is my general impression. The point I am trying to make is that if it is easy to throw the reliability of answers in doubt, how can anyone trust them? – Ian Apr 29 '12 at 1:46
  • 3
    "the problem with this is that that personal feeling has nothing to do with if that answer is correct or not". Considering that a language functions because of the collective sum of what the population determines to be correct language, then personal feeling has everything to do with if an answer is nearer to correct or not. As compared to other more objective studies, that's where personal feeling has nothing to do with correctness. For language sites, I guess the theory is that there should be a sufficiently large amount of voters to approximate a correct answer. – Flaw Apr 29 '12 at 9:07
  • 4
    I agree with Flaw. As with any SE site, peer-review (via voting and comments) is what gives credibility to the right answers. Moreover, we're lucky enough to have dedicated native-speakers like sawa and Tsuyoshi Ito (and many others) who call out errors and matters of unnatural usage that may escape the attention of some non-native users (like me). Personally in this case I edit my answer to remove info that could be misleading. Referencing a variety of reliable sources is essential to answer some questions, but some really are going to come down to the "feeling" of the majority of people. – ジョン Apr 29 '12 at 14:09
  • This is a good and current example of the above process in action. Although my answer came first and was marked correct, Hyperworm's subsequent answer was superior and this is reflected in the votes it has received. – ジョン Apr 29 '12 at 14:29
  • 1
    @Flaw The key difference here however is most people on this site are not native speakers and are probably not even advanced students. So how can we say the community is qualified to judge the correctness? – Ian Apr 30 '12 at 3:56
  • Perhaps, but a glance at the most active users on the site shows that most or all are either native speakers or advanced students. Maybe it would help illustrate the problem if you could point out some examples of incorrect information here that has gone unchallenged? – ジョン Apr 30 '12 at 14:37
  • @ジョン This question was started because I had seen at least 3 examples of very incorrect answers that had been voted up. I don't want to provide them as examples because in a such small community I think it would be counter-productive to "name names" if you will. I would hate anyone to be discouraged simply because others had incorrectly up voted their answers. – Ian May 1 '12 at 0:52
  • @Ian You're right - that wasn't a good suggestion on my part. – ジョン May 1 '12 at 2:35
  • 6
    Have you corrected it? Pointed out the issue? Is it serious, or are we picking nits? Or do you prefer to simply ask pointless questions on meta? – jkerian May 1 '12 at 5:47
6

JLU is receiving somewhere in the ballpark of 500 visits per day. Roughly speaking, we're getting about 40 upvotes per day, spread across questions and answers(EDIT: vote estimate corrected). This indicates that the vast majority of the visitors are simply reading and commenting, and witholding votes from situations... possibly because many of them don't fully understand.

Rather than suggest that you're underestimating the JLU-ians ability to identify a correct answer, I'd argue that you're vastly overestimating the ability of SO-ians to do so. Simply because a piece of code can be tested, does not mean it's good quality, standards-conformant or reliable. I've seen examples of literal stack overflow tricks being marked as 'solutions' on the main site. This is not a problem unique to JLU in any way, shape, or form.

  • just looking at the stats for this month (it's the 30th so it'll reset tomorrow BTW) there were 29 voting users; 1301 total votes (44 per voter); 43 votes a day (1.5 votes a day per voter). Some voters vote a lot more than others though, and most of the voters are pretty active members I think. – cypher Apr 30 '12 at 9:39
  • @cypher Also remember some of those votes are for questions. Probably 25-40% given the average question score. – Troyen Apr 30 '12 at 17:16
  • Yeah, those numbers also include downvotes as well. You can get the ratio of upvotes/downvotes and questions/answers of voters from individual profile pages, but AFAIK you can't get those numbers over a specific time period. – cypher Apr 30 '12 at 22:57
  • Even if I am underestimating the ability of people on SO to select a good answer, I think you are vastly over estimating the ability of JLU users to not select a flat out wrong or very incomplete answer. I don't intend that as an insult to JLU users, I simply think it is the natural result of having a voting system that is largely opinion based. – Ian May 1 '12 at 0:48
  • 3
    You missed my point completely. The same is true on the other sites you mentioned. Are you afraid that your correction will be ignored... or... what exactly? You seem to be complaining that the answers here aren't going to have textbook-level authority, which is not, frankly, a reasonable complaint. – jkerian May 1 '12 at 6:09
6

I'm a bit confused as to where this idea of masses of incorrect information on JLU comes from. Misinformation definitely crops up, but generally I see it dealt with very quickly.

Keeping in mind the obvious proviso that one shouldn't unquestioningly accept info on a community-moderated site, I feel that the reliability of the info here is higher than other communities/forums I've seen. This is probably due to a combination of the SE Q&A format, and the number of knowledgeable and dedicated contributors.

Time that could be spent discussing perceived differences between JLU and other SE sites would (IMO) be better off spent embracing the community-moderated format, and correcting the mistakes as you see them.

  • 3
    +1 "the obvious proviso that one shouldn't unquestioningly accept info". This. Just this. – Flaw May 2 '12 at 6:21
-4

Accept rates shouldn't be shown on this web site.

They're useful in Stack Overflow for two reasons:

  1. Some users don't know about accepting answers, and need to have it explained to them. On JL&U, an alternative way of detecting this problem would be seeing if a question asker has the "Scholar" badge.
  2. Some users are inconsiderate, and post drive-by questions and don't provide any feedback on answers. Although JL&U has had problematic question askers, I don't think this has been a problem.
  • 2
    I'm confused as to how accept rates is related to the topic at hand. – atlantiza May 3 '12 at 3:16
  • @atlantiza If the OP can't tell whether an answer is correct or not, then they shouldn't be pressured into accepting answers by the accept rate statistic being displayed. – Andrew Grimm May 3 '12 at 3:40
  • 1
    That said, you could say the same for most SE sites. If someone really cared about their accept rate then in this case their best bet would be to follow up on the answers with some research of their own, take the votes and comments of each answer (and perhaps the answerer's background) into account and mark the one they are most satisfied with as correct. – ジョン May 3 '12 at 19:18
  • 1
    I agree that this could apply to other SE sites. And, I'd like to add, if they aren't satisfied with the answers, they should leave a comment saying what more they are looking for. – atlantiza May 3 '12 at 20:16

You must log in to answer this question.

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