11

One of the things that we occasionally see on P.SE (and more often at the start of a new college semester) is new users posting their homework. With single posts, we are often able to comment and close before the user goes too far down the "this isn't what the site is about."

However, it isn't rare (or at least as I perceive it) for these homework problems to be in rapid succession from the same user.

So, the stats I'm wondering about:

For users with less than 125 reputation at the time of posting the questions (the rate limit threshold), who post their first two questions within 90 minutes of each other:

  • What is the eventual outcome of these posts?
  • What percentage of the users end up question banned? or severely rate limited?
  • Just curious: how many of them read the tour page (getting the informed badge) before posting a question?
  • How does this compare to users with less than 125 reputation who post their first two questions separated by a day or more?

The driving force on this is that these posts take up much more community moderation time (another set of close votes, additional comments explaining) and end up being a poor experience for the new user on the site. Getting an idea of how often this happens would aide in letting us create better proposals for how to address this and maybe help raise the awareness of the new user experience on the site at various levels.

Queries should identify the following pairs of posts (all 10k links):

  • Interesting. Some of the queries may be challenging or impossible to get from a single SEDE query. I might give this an attempt though. – maple_shaft Feb 4 '16 at 13:14
  • 2
    Good question, I've noticed that there appears to be a large number of "easy" questions being asked by people who just joined. Makes me wonder if students are creating multiple accounts to ask more questions. – Tom Penny Feb 4 '16 at 13:15
  • @maple_shaft there is a PostsWithDeleted table to work from, though I'd have to dig into it to see how much information is retained to be able to construct such a query. There is also the difficulty where the user has since been deleted (automatic process deleting users who have no posts) though some assumptions might be able to be made on those. – user40980 Feb 4 '16 at 14:17
  • 1
    @TomPenny I suspect not. The extreme size of the "new student" pool that is constantly being refreshed is likely enough to be able to describe the influx of 'easy' (or poorly researched) questions by people who have just joined. – user40980 Feb 4 '16 at 14:18
  • I gave this a second look and don't believe it is possible to query this with SEDE based on your SPECIFIC criteria, at least not without rather extreme measures. The fact is that we can't easily say what the Original Question Owner's reputation was at the time of the question. Reputation is tracked at the User level and is not tracked over a measure of time. We could probably look at when the user was created within a certain threshold of time? – maple_shaft Feb 5 '16 at 19:32
  • @maple_shaft Looking at the first two questions of users created in 2015 would probably put it on the right track. One could also remove users who have positively scored answers that were posted prior to the first question post as being users likely over 125 rep. – user40980 Feb 6 '16 at 0:52
  • 1
    SE team folks have no time to find stats for this. They are very busy now, collecting other, very important stats, on how suspensions at one site might impact elections on another – gnat Feb 6 '16 at 8:19
  • You cannot do this with Data.SE because the Deleted posts do not include the OwnerUserId of who posted the question. A quick look at an old new users query I had saved shows that P.SE had about 39,943 new user accounts in 2015, with 3675 containing 1 or more non-deleted posts, and 761 of those being more than 1 non-deleted posts. – Rachel Feb 9 '16 at 20:49

You must log in to answer this question.