Two weeks ago, we cancelled the Moderator election for this site on the grounds that there were insufficient candidates. I posted this meta post to investigate the source of the problem and determine when, if at all, should the election be rebooted. The responses mostly came in three forms.

  1. There was a lack of awareness that there was an election. This is both folks saying that they weren't actually aware, or that they figured that others who might run were not aware. We will look into things which can help improve awareness if the election is restarted. That said, Meta is an essential part of governance and moderation on a site. With several posts on the topic of the election (the initial announcement, two separate posts on the Q&A, and the extension announcement itself), a separate concern comes up with the lack of overlap in “People who might be interested in moderating” and “People who pay attention to Meta and the status of the election”. Which brings us to...

  2. The people who are standing up now, are only hesitantly doing so on account of the shortage of candidates. While a hesitant candidate is better than no candidate, it doesn't paint a good picture if the only people standing up are those reluctantly doing so because not enough people are motivated to the task.

  3. There's people who actively and frontforwardfacingly are disinterested in the prospect. Two such statements were put forth, and it concerns me more because these are relatively high rep users. One noted that there's too much overlap with Cross Validated, while the other finds the current quality of the site to be too poor for interest in moderation.

Dawny33, one of the candidates who stood up during the initial nomination phase, had this to say in response to one of the posts.

And, this is exactly why it needs more users to be involved in moderation and helping the site, as it grew in volume at a very impressive rate.

But we don't just need users to be merely involved in the moderation of the site - we need users to have a vested interest in being involved. And I can't say that we're particularly seeing that here. By the volume of question traffic here, we certainly have a topic space - that's what helped qualify you folks for graduation in the first place. But that topic space is as needing of moderators as it is of experts to answer questions. Absent that, it may well be better off folded into one of the larger sites.

So this is our call for action here. A site isn't held atop the pillar of a couple individuals - there may only be three moderators to be elected in an initial election, but moderation is a job distributed across the community. I don't want to see people stepping up sheepishly suggesting that they might have enough time to spare. I want to see that there are people who don't just care about this topic but care about this site and making it work. And I want to see more of these people. If that’s you, then now’s your chance to speak up, step up, and start moderating: voting, flagging, editing, identifying and discussing thorny issues when they present themselves. Demonstrate that you have the community that can support the high volume question base that you've accumulated, and stand on your own as a site.

  • 2
    $\begingroup$ are there any metrics available about the level of moderation on this site compared to others? the recent election on crossvalidated wasn't overrun with nominations either (of course they already have a bunch of moderators), so I'm just wondering how big the differences are... (btw one of the nominees' goal actually was to terminate this site because of the overlap...) $\endgroup$
    – oW_
    Aug 29, 2017 at 20:53
  • 2
    $\begingroup$ I left that comment because I haven't seen that user ever in the review queues here. So, if you don't contribute to moderation yourself, you don't get to complain about poor formatting and post quality. I know I am being blunt, but that's how community works. $\endgroup$
    – Dawny33
    Aug 30, 2017 at 6:35
  • 2
    $\begingroup$ Clearly any path you choose that includes keeping the site, needs to have Dawny33 as a moderator. $\endgroup$
    – Stephen Rauch Mod
    Aug 30, 2017 at 15:27
  • 3
    $\begingroup$ @Dawny33 I do actually think people get to complain about it, if it ends up being the reason that they don't contribute to moderation themselves. Which is what I'm concerned about being an actual thing here and what I want to see the site rise above. You, I see doing a whole lot of work and enthusiasm on the prospect, and that's why I specifically quoted you as having stood up during the initial nominations. But no small part of the reason we need mods is because the current situation is the site standing on just a single devoted mind. I quoted you because I want more people to join in. $\endgroup$
    – Grace Note StaffMod
    Aug 31, 2017 at 15:57

3 Answers 3


It all makes sense. In practice, it's odd that we're apparently short of moderator activity, had at least one user (Dawny33) with clear willingness and track record to become a moderator, but can't make this person a moderator.

It looks like a real election is premature, but, can't we swap out/in interim moderators for now? I know we did it before. Is there a process or criteria for that?

All the more because I'd like to pass on the moderator diamond for now, and I think I do most of the moderation.

  • 1
    $\begingroup$ The original plan was to use the election to handle the moderator situation. Given that... not so working, we'll be reaching out to you privately to discuss things further. $\endgroup$
    – Grace Note StaffMod
    Aug 29, 2017 at 19:06

Two such statements were put forth, and it concerns me more because these are relatively high rep users. One noted that there's too much overlap with Cross Validated, while the other finds the current quality of the site to be too poor for interest in moderation.

Firstly, DataScience and CrossValidated have very clear and separate scopes for each other. So, converging them is a very bad idea. Stats is a part of data science, so the overlap is justified. But, saying DataScience site should be merged into CV is like saying DataScience is nothing but stats. It is not. There are domains like Data Engineering, etc which are not on-topic on CV[explained clearly in the answers linked in the preceding text].

One cannot expect a spotlessly clean quality of posts from a newly graduated site. It takes time, but this site is constantly improving, thanks to all the users in the reviews queue. The user who complained about poor quality hasn't been clear about what they find lacking in the quality. If they can do it, I can also go ahead and address that in detail. + comparing the quality on this site to sites like SO (where there are scores of 20k users) is not justified.

I am not sure about the lack of interest amongst users in stepping up to moderating this site. This happened in the recent election of CrossValidated too. The users who stood up, stood up so that the election doesn't get called off.

I can only speak for myself. I come here almost every day asking and answering questions, and I like the helpful response to my questions. I was involved with this site for a long time and enjoyed that time too, so that is why I stepped up.

  • $\begingroup$ I think it's the sheer extent of the overlap that concerns people. What topics can I ask about here? gives five examples of question titles, four of which sound perfectly appropriate for Cross Validated. If it's expected that the body of such questions should focus on particular facets of the problem - Data Engineering, implementation & tools, or whatever - that would be off-topic on CV, shouldn't that be spelt out? Especially as Data Science is a rather new field & many people won't have a clear notion of what it concerns itself with. $\endgroup$ Sep 4, 2017 at 8:56
  • 1
    $\begingroup$ Adding up to what's been said, it's hard to agree that the scopes of DS and CV are well outlined, even with the given answers in Meta. Although I could be mistaken, the criteria that seem to grab the most attention are (1) questions that a data scientist would have that can be answered by experts in this field; (2) questions with a more "practical" approach to machine learning than Cross Validated, even though the latter claims to accept most of them. $\endgroup$
    – E_net4
    Sep 4, 2017 at 10:07
  • $\begingroup$ It surprises me when I hear it suggested that CV frowns on practical or applied questions, or that it welcomes only academic or theoretical ones; given the great number that arise from issues encountered in the course of data analysis being carried out in biology, engineering, marketing, & a host of other areas. Perhaps it's a matter of perspective: for a programmer or computer scientist, practical questions are those about writing code in a particular language. At CV we close (or migrate, when possible) questions that ... $\endgroup$ Sep 4, 2017 at 13:10
  • $\begingroup$ ... "focus on programming, debugging, or performing routine operations within a statistical computing platform" (I'm quoting from the close-reason text). Nevertheless plenty of good, on-topic questions involve code or the use of statistical software - asking for help with the explanation of puzzling results, the interpretation of outputs, &c. And many good answers are illustrated with code examples. $\endgroup$ Sep 4, 2017 at 13:17
  • $\begingroup$ Hi @Scortchi. Nice seeing you here. The What topics can I ask about here? page is heavily outdated :( Gotto update that! $\endgroup$
    – Dawny33
    Sep 4, 2017 at 17:02
  • 1
    $\begingroup$ Hi @Dawny33: Congratulations on becoming a moderator! Glad to see there's a Meta question on this now. $\endgroup$ Sep 22, 2017 at 13:27

I understand the comments about low quality questions on this site. They increase the effort to moderate the site and make it seem to people browsing the site for the first time that there is not a lot of value here. This prevents potentially interested users (who might be active on crossvalidate or stackoverflow) to participate here and contribute, which in turn prevents faster improvement of the overall quality of the site.

Maybe something can be done about this upfront that doesn't involve any moderation. When you start formulating a question the only guidance you receive is this:

How to Ask

Is your question about data science?

We prefer questions that can be answered, not just discussed.

Provide details. Share your research.

This is clearly doesn't seem to do the job. How to ask a good question on data science SE still doesn't have a satisfactory answer. If we can provide some more details what makes a good question this could help everyone.

This should include at least something like if your problem involves a specific dataset, then provide a sufficiently detailed description and/or a sample of it, etc.

On a different note, if the metrics looked good enough for graduation, then I think it's fair to assume that this site will gain further traction and improve in the long-run. If not, maybe the metrics aren't that great...


You must log in to answer this question.

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