10

I was very worried when we first decided to allow questions be migrated here from ELU.

Now that it's been several months, how is it going?

I was prompted to ask this question because three questions were just migrated from ELU and at least one of them doesn't seem to follow the very clear "Don't Migrate Crap" rule.

the difference between "come of" and "come from"

  • Has horrid formatting, terrible punctuation and bad capitalization.

I know that we aren't expecting perfection from questions asked here... but ELU needs to be editing these questions for at least basic quality before they are shunted to ELL.

At this point, 9% of the questions migrated here in the last 90 days were rejected... that's nearly one in ten... and I have a feeling that this number is lower than it should be, since many of the questions have a score of zero, which means that they're likely not very good questions.

Here's a couple of examples:

Which one is correct to use?

This question has been here for over a week and didn't even attract any comments from regular users to ask for clarifications, the two answers don't have any votes, either... the OP hasn't created an account here...

Do the users here tend to ignore the migrated questions? This type of question usually attracts comments... or votes... something. It's utterly mystifying that this one has no votes.

Why did the writer use "would" instead of "loved"

This question has a downvote as does the answer from a regular user... and it's still open despite not being specific about what part of the sentence is in question... should it not have been closed as needing detail or proofreading? Again, the user hasn't come here and created an account and hasn't responded to requests for clarification.


Here are the 90 day migration stats for anyone under 10K:

90 day migration stats

Is there anything to do about this or should we do anything about this?

Are we OK with having these questions moved here, regardless of any effort on ELUs part to improve them or make sure that they're actually good fits for ELL?

If not, should we make a request to ELU that they address this? Perhaps before shutting down the pipe entirely we can ask mods to take a more active part in reviewing the migrations or address it in their meta?

If so, how can we make users more enthusiastic about actually addressing them - voting, commenting, answering...

  • 1
    Every question ever migrated here from EL&U (except the handful that got deleted) – Shog9 May 12 '16 at 3:34
  • Related discussion: meta.english.stackexchange.com/questions/7784/… – snailboat May 12 '16 at 3:47
  • 1
    @snailplane Thanks for that... that's quite an enlightening... vision of the issue. I admit that none of my examples are quite that extreme but clearly there's something going wrong. – Catija May 12 '16 at 3:53
  • @Shog9 Is it possible to get that with a third column for "status" - ie open/closed/duped? – Catija May 12 '16 at 3:54
  • Oh, fine: data.stackexchange.com/ell/query/485079 – Shog9 May 12 '16 at 4:11
  • @Shog9 It would be useful to see question score in the query - that way we could see if it's just a few egregious examples that stick in our minds and most of the migrated questions are actually received well. I would do it but I'm already up too late. We should also keep in mind that closing a migrated question as a duplicate should be excluded from the statistics. I don't think it's reasonable to expect EL&U reviewers to search for a duplicate on ELL before they migrate a decent quality question. – ColleenV May 12 '16 at 4:26
  • 4
    W/ Score, @Colleen: data.stackexchange.com/ell/query/485081 – Shog9 May 12 '16 at 4:28
  • It looks like about 1/3 of the migrated questions have a score under 1: data.stackexchange.com/ell/query/485239/… I didn't limite it to negative scores, because often folks don't downvote questions they're closing. One the whole we're getting more good migrations than bad. I'll try fiddling with the queries a bit more later, but right now I'm going to be late for work if I don't get my butt moving. – ColleenV May 12 '16 at 13:12
  • @ColleenV if you haven't already, could you focus on migrations after the path was created? – Catija May 12 '16 at 13:32
  • Yes, that was the extra fiddling I was planning. Before the path it was moderator migrations, which were probably all fairly good quality. – ColleenV May 12 '16 at 15:08
  • @ColleenV you might be surprised. I remember going to the ELU chat at least once to complain about it. – Catija May 12 '16 at 15:10
  • I've updated to put in the earliest migration date as 09/22/2015 and added a max score parameter: data.stackexchange.com/ell/query/485239/… Out of 479 total, 17 have negative scores, and 206 have scores of 0 or less. I'm going to have to do a little more work to select only the zero rated that were closed and to filter out the closed as dupes. I modified this query to match the date range: data.stackexchange.com/ell/query/485285/… to give some insight into close reasons. – ColleenV May 12 '16 at 15:58
  • Thanks for doing all of this work! I'm rubbish with SEDE but I'm glad someone here knows what they're doing. – Catija May 12 '16 at 16:00
  • I know just enough SQL to be dangerous, but not enough to be able to get insights quickly ;) – ColleenV May 12 '16 at 16:02
  • 1
    Looking at the data, I'm not seeing a big problem. It's annoying when someone goes through all the trouble to review something, then tosses it over the fence even though it's extremely low quality, but shutting off the migration path would actually lose us quite a few good questions – ColleenV May 12 '16 at 22:06
9

I've always seen this as a two-pronged problem. The first problem is the quality of the question; the second problem is the question being asked in the wrong place.

When a new user (rep of 1) asks an ELL question on ELU, I don't have a problem with it being migrated here, even if it's a bit on the "crappy" side. I understand that "migrating crap" is discouraged, but, in this case, we are often dealing with folks who have very limited English skills. I don't think the onus of training learners in the art of asking a well-researched SE question needs to fall on ELU, only for those mentors to tell them afterward, "Now that you've got that figured out, you should ask this on ELL."

I would rather get the learner plugged into the right community first, and then handle the problem of tutoring new users on customary SE standards. Frankly, if we are rejecting less than one in ten of the migrations coming here, I think we're doing pretty good.

As for this:

many of the questions have a score of zero, which means that they're likely not very good questions

Perhaps that's true, but we get a good many questions here that end up with a score of zero, whether they are migrated here or not. (Roughly 6,000 of our 25,000 questions currently sit at 0.) Question quality may be a problem, but I don't think it's a problem because of migrations.

That all said, if a question could use some formatting improvements before it gets migrated over, that would be a nice gesture on the ELU side, and much appreciated over here.

Lastly, if a learner's question is simply so bad that an ELU voter can't vote to migrate in good conscience, it would be nice if they left a comment that doesn't just point out the shortcomings of the question, but also points the new user in our direction. But maybe I'm being too idealistic on that one.

  • 1
    I don't think a score of zero is an indicator of a 'bad' question. Zero up-vote questions are most likely candidates for editing to make the question clearer, or they're just normal questions that no-one feels are particularly good or particularly bad. I don't like that migration is part of the close process. It would be nice if it were a more positive process. "I think you would get better answers on site X" is a different mindset from "Your question doesn't belong here, maybe site X will tolerate it." (I know that's not how it's supposed to be, but that seems like how it is) – ColleenV May 12 '16 at 21:23
  • 1
    @ColleenV - I view migration as somewhere in between those two interpretations (i.e., something like, "I think your question would be a better fit on site X"). As for what zero represents, I think you're on-the-mark; it's not so much that the question is "bad," it's just not "very good." – J.R. May 12 '16 at 21:47
7

I wonder if the angle I'm looking at this from is wrong.

The "golden rule" of migration is definitely "Don't migrate crap"... but I think that what we really need to understand is:

What factors are the users of ELU using to determine that a question should be migrated here rather than simply being closed on ELU.

I believe that, if only the mods were migrating here, the quantity of migrations would be significantly reduced and that more questions would be simply closed rather than being sent here.


For example, this question was just migrated here:

Different versions of "I have no question of my grandmother's dedication", which are acceptable?

The entire content of the question is:

"I have no question of my grandmother's dedication"
"I have no question of whether my grandmother is dedicated"
"I do not question my grandmother's dedication."

Which can be used, grammatically?

Why is this off topic for ELU?

Nothing overt about the grammar or spelling or punctuation implies that this question is from a learner.

It wasn't well formatted before I added the quote text and it doesn't explain what the OP is trying to understand... but what about this question is a bad fit for ELU?

All of these issues are a reason to either downvote for quality reasons or close as "unclear"... they don't seem to make the question "off topic".


Here's another one:

In a comment on this question, Alan Carmack pointed out this question:

Verb tense after someone died

It has three possible dupe targets on ELU, one of which is essentially identical and is even protected. I'm guessing that some of the close voters actually voted for the dupe... but somehow it ended up being migrated here instead.

This question is well-written, clear and very much about English Usage. How is that "off topic"?


So, while J.R. is perfectly correct in addressing the issues I mentioned in the question and the points made are valid... but what I think we may need to be asking is

Are there some users who are over-using the migration option on questions that aren't really off topic on ELU?

I think that, more than complaining about crap being migrated, this is an internal discussion that ELU needs to have to decide if there's a way to encourage users to work to improve and keep questions that don't actually fall outside the ELU scope.

  • 4
    See also Shog's blog post, especially this bit: "Be a bit jealous of your site - don't blithely turn askers away simply because their question could be asked somewhere else. Don’t hit them over the head with your scope, help them tailor their question to fit into it - and if that means your site’s scope overlaps a bit with another site’s, so be it." Seems apropos. – Nathan Tuggy May 14 '16 at 6:25
  • 1
    Yes! "Get a better reception" is not a reason to migrate. Questions need to actually be off topic to move. Questions about Star Wars might get better answers on SFF than M&TV but they are not off topic so moving them is not appropriate unless the OP changes their mind and asks to have the question moved. – Catija May 14 '16 at 13:54
  • I think you bring up a good point not only for EL&U, but for ELL as well now that Language Learning is in beta. ELL has accepted questions about teaching English in the past (someone set me straight about that a while back when I tried to close a pedagogy question), but now I'm starting to see questions getting directed to the other site and close voted. I think you can make askers aware of another site without necessarily closing their question as off-topic (or migrating it if there is a migration path). – ColleenV May 15 '16 at 12:29
  • So, it seems that the fact that ELU has a close topic for "Does not show research" listed under "off topic" has created some confusion about what "off topic" actually means... And, of course, this relates to the issue of the fact that all custom close reasons are listed under "off topic" regardless of whether that's actually the case or not. – Catija May 15 '16 at 19:11
2

Here are links to various queries from the comments - I've made this a community wiki so we can share them more easily. If you come across or write a query that helps give some insight into migrated questions, please add it.

  1. Every question ever migrated from ELU with closed reason and score from Shog9 https://data.stackexchange.com/ell/query/485081

  2. Questions migrated from ELU since 09/22/2015 with a parameter to select the maximum score (shamelessly forked from Shog9's work): https://data.stackexchange.com/ell/query/485239/migrated-questions-from-elu-with-scores-less-than-parameter?MaximumScore=50

    • This query shows on 5/12 that only 17 out of 479 migrations were down-voted below zero, with the most negative being -3 (+4/-1).
  3. Counts of migrated questions from all sources, how many were closed, and the close reasons: https://data.stackexchange.com/ell/query/485285/migration-quality-control

    • This query shows on 5/12 that only 36 out of 479 migrations were closed for reasons other than 'duplicate'. That's about 7.5%.
  4. Here is a query that lets you set the minimum and maximum score for questions migrated since the migration path opened (09/22/2015): https://data.stackexchange.com/ell/query/485239/migrated-questions-from-elu-with-scores-less-than-parameter?MaximumScore=50&MinimumScore=4

    • This query shows on 5/12 that 43 of the migrated questions have a score of 4 or higher, with the highest two having a score of 39 (+43/-4) and 25.
  • that first one doesn't actually show the closed reason other than dupes... I looked through the first few months of them and (were it to be believed) no questions have been closed other than as dupes. I think that the rejection of these questions is preventing them from appearing there. – Catija Jun 12 '17 at 16:56
  • @Catija I think many of the rejections have been roomba'ed. We probably need multiple queries to get the whole picture. The rejection rate was 7% the last time I looked at the 10K tools, which I don't think is terrible, considering the number of questions that were well received. We just have to kiss a lot of frogs to find the princes. – ColleenV Jun 12 '17 at 17:12
  • That's low. It's been 14-15%... though, if there are a lot of 0 score questions with no answer, they may have simply slipped through the cracks. – Catija Jun 12 '17 at 17:14
  • @Catija If you look at query 3, you should get a better idea of the close reasons. We could close a lot more questions than we do, but I think it's better to try to salvage them. Many questions with very useful and interesting answers started out pretty bad. What would be interesting is to see what % of "bad" questions the migrations make up of all "bad" questions in the last 90 days. – ColleenV Jun 12 '17 at 17:24

You must log in to answer this question.

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