6

We currently do not require a person to register with the site in order to ask a question. Should we change this so registering is required?

6 Answers 6

6

Yes!

I find that registration increases the likelihood of responding to comments, accepting answers, putting effort into the questions and such. If a user makes zero commitment to our site, I don't see how we owe them an answer. The small extra hurdle of registration is likely to make new users think twice before asking, and that's good: forcing people to put a little more thought into their first questions is a step in the right direction. Requiring registration doesn't solve all problems, but I think it's for the best. New users will continue to be welcome here, but their integration could be enhanced.

A crucial benefit of registration is eased communication. The user can accept an answer, receive notifications of messages, be suspended (in the very rare cases where it's needed) and such. They can also benefit from the reputation points, which makes their use of the site smoother and somewhat reduces the workload of moderators.

Lastly, I should add that requiring registration doesn't make asking questions much harder — many websites require registration and a policy like that here would not stand out. I absolutely don't want to close off the site and keep it exclusively to old users. I just want new users to commit to interacting with their question even if they only have one — it also helps them give what they need, as they often struggle to formulate questions at first.

If the community gives its support, we mods can figure out how to go about this. Some SE sites do require this, so we should be able to do the same. I just checked some moderator documentation and found that this is indeed configurable.

2
  • Did this become the policy after this discussion, and was it changed recently? I've been noticing unregistered accounts asking questions in recent weeks (mainly because they don't bother to interact with the answers they get), though I'm not sure if it's just confirmation bias.
    – Cairnarvon
    Commented Feb 29 at 2:28
  • @Cairnarvon Thanks for resurrecting this. I think at the time the upvotes were low enough that a clear consensus wasn't reached, but that's not the case now. We're discussing moving this forward with StackExchange now.
    – cmw Mod
    Commented Mar 1 at 15:13
3

This is not a real answer, but rather a collection of data and some thoughts on it.

Data

Here is a quick numerical study based on the Stack Exchange Data Explorer. I have listed below the numbers of questions owned by "drive-by users" and those owned by "committed users" sorted by question score:

Score Drive-by Committed Percentage
(negative) 2 10 17
-5 1 0 100
-4 0 2 0
-3 0 1 0
-2 0 5 0
-1 1 2 33
0 2 29 6
1 12 142 8
2 18 283 6
3 28 476 6
4 19 563 3
5 13 564 2
6 15 516 3
7 6 418 1
8 3 364 1
9 7 232 3
10 3 173 2
11 1 122 1
12 1 97 1
13 0 64 0
14 1 50 2
15 0 34 0
16 0 28 0
17 0 26 0
18 2 21 9
19 0 12 0
20 0 10 0
21 0 5 0
22 0 4 0
23 0 3 0
24 0 5 0
25 0 8 0
26 0 2 0
27 0 3 0
28 0 4 0
29 0 5 0
30 0 3 0
31 0 1 0
32 0 3 0
34 0 1 0
35 0 1 0
36 0 2 0
37 0 1 0
43 0 1 0
44 0 1 0
47 0 1 0
49 0 1 0
50 0 1 0
55 0 1 0
81 0 1 0

Here "drive-by users" are those for whom the time difference between creation and last activity is at most one hour, "committed users" are those for which the difference is at least 100 hours. The last column shows the percentage of posts of the given score owned by drive-by users. The first row combines all subzero scores into one line.

The data comes from this SEDE query. (Thanks for the pointer brianpck!) If anyone can modify the query to produce the table automatically, that would be great, as it would allow tweaking the parameters with less effort.

It should be noted that many low scoring questions are deleted automatically or by their owners, so the data is not very descriptive below zero. The data contains no deleted questions, so it does not help in analyzing the questions of lowest quality. The included questions with negative score have managed to get an answer (with a positive score) before being (as such questions often are) closed. The data also says nothing about closures (which could be addressed in a separate query), but many closed questions end up deleted and wouldn't show up anyway.

Conclusion

The following rough summary table shows that the fraction of posts owned by uncommitted users goes steadily down as a function of score. The total number of nice questions (score at least 10) owned by these users is 8. This is a very small number compared to the 695 ones owned by the others.

The conclusion seems to be that uncommitted users produce significantly worse questions than committed ones as measured by the question scores. The last column is a decreasing one. The questions by the uncommitted are certainly not worthless, but I would argue that they do not substantially add to the site either. The loss of question influx from a policy like this would seem to be quite minor, and it would skew the quality distribution of questions in the right direction.

Score Percentage owned by drive-by users
…-1 17
0 6
1…3 6…8
4…6 2…3
7…10 1…3
11…20 0…21
21…10 0

1 There is an outlier at score 18. It is surrounded by zeroes in the drive-by column, so I have taken it to fall within the range 0–2. This is an expected statistical fluke having to do with small integers and not of much significance.

7
  • Great work! So we'd have prevented 4 questions with a score of 0 or less in our entire history, is that correct? I wonder what percentage of drive-by questions were asked by registered users. P.S. "Data comes, data is", hmm?
    – Cerberus Mod
    Commented Dec 21, 2021 at 19:16
  • @Cerberus Thanks! These numbers tell nothing of questions that are closed and deleted, which is a pity. That would be the most interesting kind of thing, really. Therefore I don't think your interpretation is quite correct. I wonder if there's a way to learn about the historically closed questions, many of which are deleted now... (I like to think of the English word "data" as an uncountable noun akin to "information", not typically a plural of "datum".)
    – Joonas Ilmavirta Mod
    Commented Dec 21, 2021 at 19:22
  • As a native English speaker, I still feel weird using data in the plural even though it's grammatically correct. "The data were this, the data were that".
    – Adam
    Commented Dec 21, 2021 at 19:47
  • @JoonasIlmavirta: Ah, so do you mean it says nothing about closed questions, nor about deleted ones? Or does it say nothing about questions that were closed and then deleted?
    – Cerberus Mod
    Commented Dec 21, 2021 at 20:04
  • @Cerberus Deleted questions are not included, and the data doesn't tell whether a question is closed or not. Being closed is included in the database, but someone other than me should write a new query. I edited the answer to explain that better.
    – Joonas Ilmavirta Mod
    Commented Dec 21, 2021 at 21:12
  • 1
    @Adam The scientific community is divided on the matter. I find it more natural and, more importantly, useful, to treat it as a singular. The fact that it's a plural in Latin need not mean that it's a plural in English.
    – Joonas Ilmavirta Mod
    Commented Dec 21, 2021 at 21:13
  • @JoonasIlmavirta: Thank you, I'm sorry I skipped over that part earlier in my answer. I've made it clear now that we don't know (yet...) about deleted questions.
    – Cerberus Mod
    Commented Dec 22, 2021 at 0:41
2

I have enabled the requirement to register to ask questions. Please note, though, that users can still answer questions without needing a registered account.

1

No, we should not.

The first question to ask would be what is the problem that this is supposed to be the solution to? Latin.SE is one of the best and most enjoyable SE sites I've ever been on. It has the vitally necessary core of high-knowledge users, and it also has a decent amount of participation from people like me who are at lower levels of skill. I've never seen any kooks. I haven't noticed any users who pose as high-skill users but who actually post nonsense. The site does not seem to be exploited by students who want to get their homework done for them (probably because almost nobody these days takes a class in Latin or Greek except by choice). Almost all SE sites would or should wish themselves to be doing as good a job as latin.SE.

If the site's biggest problem is people who want their tattoo ideas written in perfect Latin -- wow, what a great problem to have. (Well, it's unfortunate that there appears to be nobody who wants a tattoo composed in perfectly idiomatic faux Homeric Greek, in which case I would be happy to lend my crappy non-expertise.)

Forcing registration has clear disadvantages. It turns away newbies who are not sure they want to participate. For a site this small, newbies are like lumps of gold the size of a bowling ball. Nothing should be done that might keep them from coming.

Joonas Ilmavirta says:

[Users who register] can also benefit from the reputation points, which makes their use of the site smoother and somewhat reduces the workload of moderators.

By the way, thank you for your work as a moderator. I realize that it's a tedious and thankless job. But if you're finding that the workload is a problem for you, a really good strategy is just to do less. There are other mods. Lack of super-active moderation typically doesn't create any problem. The SE system allows the hoi polloi to take care of most things simply by commenting and downvoting.

And reputation points are not a benefit. They are a gamification strategy used to lock in users, in much the same way that pernicious social media monsters like facebook seek to lock in users. I doubt that anyone here derives gratification from looking at their rep total and seeing that it's a certain number. Seriously, I get gratification from things like walking my dogs or having a zoom call with my mom.

I very frequently ask questions anonymously on SE sites. I find that it works better for me that way. Not having an account prevents me from getting caught up in the destructive psychological cycle of desiring gratification from upvotes. There are indeed SE sites where I don't participate, specifically because they require an account to ask questions or post answers.

9
  • 1
    "The site does not seem to be exploited by students who want to get their homework done for them" – I have seen this, but it is seems to be rare. What we see more often is translation requests without enough context; other users were annoyed by these, which is why the policy was changed such those requests without enough context are closed. So I do think there is a legitimate dilemma, where a compromise must be reached between the interests of experienced users ("we don't like the site if it gets mainly these uninteresting types of questions") and new users ("I'd like my phrase translated").
    – Cerberus Mod
    Commented Dec 20, 2021 at 20:08
  • 1
    "And reputation points are not a benefit." – But they are! I don't mean that anyone should care about the points per se, but reputation brings privileges that allow you to vote, edit, comment, and generally participate more freely. This is a designed feature of SE sites, making the site's use easier for the more active or reputable users.
    – Joonas Ilmavirta Mod
    Commented Dec 20, 2021 at 20:45
  • 1
    One additional point that caught my eye (but not to necessarily jump into a comment train) is that some users who might otherwise participate see too many tattoo requests and think that we're a non-serious endeavor, and therefore they decide to not participate. I know this is a real thing because I've had fellow (now former) colleagues and academic friends who won't participate here because they don't see it rising to their level. We'll never reach all of them, but they would be "newbies", too.
    – cmw Mod
    Commented Dec 20, 2021 at 21:34
  • @cmw: How many "silly" questions are there: 1/30--less? Your colleagues must have a remarkable perception if all they can see are the unorthodox ones. Please see the debate with Joonas, above.
    – tony
    Commented Dec 21, 2021 at 12:20
  • 1
    @tony That doesn't sound too welcoming to knowledgeable users to me...
    – cmw Mod
    Commented Dec 21, 2021 at 19:01
  • @cmw: I don't understand why a handful of questions on the "Star-Wars" model (which was well-subscribed) would deter people from joining, when there are many other Qs. of high-quality academic content.
    – tony
    Commented Dec 22, 2021 at 9:36
  • 1
    @tony tag:english-to-latin-translation is one of our most popular tags for a reason. It's simply put: some users here like those questions, some do not. Some potential users do not come because of an abundance of those type of questions. The lines have been drawn, and the status quo wins out.
    – cmw Mod
    Commented Dec 23, 2021 at 4:12
  • @cmw: I question "abundance" and "lines have been drawn" is reminiscent of a division into armed-camps. Don't these potential users wish to pass-on their superiority, in a teaching function? Why not? They might even feel good about themselves. There are brilliant exponents, of Latin, on this site. Occasionally, even I have got one-or-two of these to revise a posting. This teacher-student symbiosis is invaluable, wouldn't you agree?
    – tony
    Commented Dec 23, 2021 at 9:47
  • 1
    @tony No, in the real world, teachers get paid for work they do.
    – cmw Mod
    Commented Dec 23, 2021 at 12:43
1

I'm a bit undecided, but I'll provide a "no" answer so that people might vote on it.

On the one hand, registration does all the good things mentioned by Joonas. On the other, it increases the threshold for any participation on our site, and we already get few questions as it is, a couple a day.

I personally don't have a problem with users who have little commitment to our site: I'm happy with its being completely open even to casual users.

Edit:

Some very interesting data gleaned by Joonas! I would summarise them as follows:

  • 2 non-deleted questions with a negative score were asked by drive-by users.
  • 2 〃 with a score of zero
  • 129 〃 with a positive score
    • 52 of which with a score of five or greater

What we do not know:

  • How many deleted questions were asked by unregistered and registered users?
  • What percentage of drive-by users was unregistered?
  • (What percentage of users that show up as unregistered are in fact deregistered, which is what happens when a user account is removed? This is interesting though less relevant to drive-bys.)
8
  • 1
    I personally don't have a problem with users who have little commitment to our site: I'm happy with its being completely open even to casual users. Well said. Latin and Greek are already sufficiently hard and forbidding without erecting additional barriers.
    – user3597
    Commented Dec 20, 2021 at 19:46
  • 3
    It would be interesting to look at a list of questions by unregistered users (is there a way to do that with a search?) and check their quality. Even a few high-quality questions by unregistered users would incline me towards this answer
    – brianpck
    Commented Dec 21, 2021 at 0:12
  • @brianpck: I think Joonas is looking into whether that is possible using SE's database tool: chat.stackexchange.com/transcript/message/59936478#59936478
    – Cerberus Mod
    Commented Dec 21, 2021 at 0:53
  • @brianpck and Cerberus: Having looked into it a little, it seems that the Stack Exchange Data Explorer (SEDE) does allow searching for whether users are registered or not. The database is great but doesn't cover everything. Of course we could do a more manual search on the site, but one has to bear in mind that ill-received questions are deleted automatically and leave a bias.
    – Joonas Ilmavirta Mod
    Commented Dec 21, 2021 at 10:00
  • If anyone knows how to do an SEDE query for the following, I'd be happy: "Given a score range (like 5-10), what percentage of questions with that score are owned by unregistered users?" Exploring that with different ranges should give an idea of how well questions by the unregistered users actually fare.
    – Joonas Ilmavirta Mod
    Commented Dec 21, 2021 at 10:01
  • @JoonasIlmavirta This query gives something like what you're asking for: data.stackexchange.com/latin/query/1533189/…. There are a few false positives, since it looks like the closest you can get with a SEDE query is to check for "users who never came back after posting." I only found a few false positives, though.
    – brianpck
    Commented Dec 21, 2021 at 16:28
  • @brianpck Looks promising! Using the time difference between creation and last access is a clever proxy for unregistered users. That query can hopefully be tweaked to compare against more established users and see how the ownership percentage (drive-by/committed) depends on question score.
    – Joonas Ilmavirta Mod
    Commented Dec 21, 2021 at 17:36
  • @brianpck I posted a separate answer with some data.
    – Joonas Ilmavirta Mod
    Commented Dec 21, 2021 at 18:32
-1

To register onto a site is easy to do and allows the user to remain anonymous. Therefore, how does this deter so-called "drive-by" types who just want a tattoo-motto and will never be heard from again?

Is registering synonymous with commitment?

Unless there is compelling evidence, I would leave things as they are. The fly-by-night users can be ignored. Mind, motto-requests often receive answers, especially if they interesting ones e.g. "memento mori"; and, recently, the chappie with a misspelling in his tattoo--the "Serenity Prayer" one--both interesting and amusing!

10
  • I updated my answer in response to this. One important aspect is that registered users are easier to communicate with. An unregistered user is difficult to reach with comments (or suspend if the behaviour is too unruly). Registering is certainly not synonymous with commitment, but it does imply a bit of commitment.
    – Joonas Ilmavirta Mod
    Commented Dec 20, 2021 at 12:30
  • The "can be ignored" bit is worth discussing more. Having low quality content on the site is bad, even if we can all in principle ignore it. I don't want to wade through poorly worded tattoo translations questions from unresponsive users to get to what is interesting. Writing comments, closing, flagging, voting, and all that takes effort. Requiring registration would make this ignoring more automatic. I fully agree that adding the requirement wouldn't magically solve the problem, but it would make it more manageable.
    – Joonas Ilmavirta Mod
    Commented Dec 20, 2021 at 12:33
  • @Joonas llmavirta: Cerberus made the point that a registration policy may reduce the question-flow. Any thoughts on that? The "fatuous question"--oft-discussed e.g. "Star Wars" (20-upvotes). I'm waiting for someone to ask, "...Nuclear Missile in Latin?", which might present something of a challenge though easy to dismiss as "stupid". I recall a debate over a one-off ring-inscription: I wanted to use a present subjunctive; yourself, an imperfect subj., which was correct. (The present was too harsh; too peremptory.) The policy has been hitherto kind, considerate--we'll help you if we can.
    – tony
    Commented Dec 21, 2021 at 9:41
  • @Joonas llmavirta: More recently, poor quality content has been met with requests for context and explanations, pending closedown--fair enough. Now, it's a registration system. In any public forum there will be good-and-bad contributions. The (apparently) silly Qs., mottoes & tattoos can generate good grammatical debate. Therefore let us facilitate a smooth, unhindered question-flow.
    – tony
    Commented Dec 21, 2021 at 9:51
  • The aim is to remain kind and considerate! I'm not advocating anything against that, and I never have. Throughout the history of the site we have rejected some questions (and in extreme cases some users) and will continue to do so, and that rejection must be done kindly too. // I agree that a registration policy would somewhat reduce the influx of questions, but I don't think the reduction would be great. I think the reduction would be mostly aimed at questions of the lowest effort. Unfortunately there seems to be no easy way to research how questions by unregistered users fare.
    – Joonas Ilmavirta Mod
    Commented Dec 21, 2021 at 9:51
  • See cmw's comment under Ben's answer for a good reason keep higher quality standards and reject some types of questions. We need to strike a balance between new and old users, and academic and non-academic users (these are two different balances, not one). I find all extremes to be detrimental, but I don't know where exactly the perfect balance is. How academic is this site now and how academic do we want it to be? I'm pretty sure there are very different opinions on this among our users.
    – Joonas Ilmavirta Mod
    Commented Dec 21, 2021 at 9:56
  • @Joonas llmavirta: Thanks. I've commented to cmw. Academic? Are you joking? Did you read Q: latin.stackexchange.com/q/17138/1982? Were you aware of the concept of "interlacing"? I wasn't. It has no grammatical parallel, in English, takes some understanding and, I'm guessing, by virtue of these, is Latin at its most difficult (academic). I intend to recommend Kinghorsey's answer for "Answwer-of-the Year, 2021". The example from "de Finibus" was in Harm-Pinkster's notes. Yes, that Harm Pinkster: you can't get more academic than that!
    – tony
    Commented Dec 21, 2021 at 12:34
  • @Joonas llmavirta: Having said all of that, there is still a place for "the little people" with their (possibly) silly Qs., translations of which could throw-up any number of Latin difficulties & debates.
    – tony
    Commented Dec 21, 2021 at 12:37
  • If you want to continue this discussion further, the chat will be more suitable, but I'll comment on your question: I'm absolutely not joking at all about whether we're academic or not. There most certainly is some deeply academic content, but that doesn't make up all or even most of our content. Some would prefer us to be more exclusively academic, while others would prefer to make even more room than now for the non-academic content. This is not a trivial balancing issue, as we'll have a broad distribution of content types anyway.
    – Joonas Ilmavirta Mod
    Commented Dec 21, 2021 at 12:49
  • Let us continue this discussion in chat.
    – tony
    Commented Dec 21, 2021 at 12:50

You must log in to answer this question.

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