2
$\begingroup$

I'm looking to put together a list of free sites that offer courses / lessons that would be useful to me and this site's audience. I don't want to do this only for the question to be closed for no other reason than it's a community effort, so before I begin, would this be a good idea?

https://datascience.stackexchange.com/questions/1011/what-data-science-courses-lessons-are-there

$\endgroup$
4
$\begingroup$

This is not the type of content that's considered appropriate for the Q&A format of Stack Exchange sites. That entire discussion is an excellent read; the question's asked very well and the answers will give you a lot of insight into both the community's position on your questions and the rationale behind it, which has to do with a number of practical considerations (as opposed to pure opinion).

Here's another relevant discussion on meta.SE that provides a counter opinion, at least to some degree. As Rachel notes, canonical list questions do exist on some sites - but in looking for examples, I haven't yet found any of third party resources except for one FAQ post on meta.hermeneutics.

Personally, I think tag wikis are the most appropriate place to collect third-party links and resources -- it says "wiki" right in the name, after all. Tag wikis are designed to hold useful information and aren't held to the same standards as Q&A on either main or meta sites. But there are a few obstacles to that approach:

  1. What tag do you put it in? If it's a coherent set of resources, it should be stored in a central location. (In our case this isn't that big a problem since we have an tag already in use that could serve this purpose.)
  2. It's generally accepted that very few users ever read tag wikis -- and there's no great way around this except to refer to the resource as frequently as you can justify in the course of other posts and comments.
  3. Users need a lot more reputation to freely edit tag wikis than they would need to edit a community wiki answer.
  4. Maintenance and outreach are a ton of work. This is an obstacle for any approach to collecting a list of resources. You need a core group that's willing to keep links working, add new resources as they become available and update/remove old ones as they change or disappear.

If you're really committed to making this work, I would advise you to start building your resource list in the tag wiki, and use this meta discussion (or a fresh one) as a place to organize your efforts. There is a long tradition of organizing things like retagging and queue-reduction efforts on meta, so such a post would be legitimate, and also give you an idea of whether this community is ready to take on the effort you're suggesting (remember, it's going to also put demands on reviewers to approve the suggested wiki edits) at this point in the public beta.


Time has made a liar of me; today I came across this list of C++ reference books on Stack Overflow. This is a great example of how much goes into doing this sort of thing right. Some important things to note:

  • Between the question and answer, these wikis have been edited over 100 times by over 50 different users. Even if we take into account that this is over a period of nearly six years, that's a lot of continuing upkeep for one question.
  • It's tagged FAQ, which suggests that there's an identified demand that this resource satisfies.
  • It's double-wiki and under collaborative lock to make sure that the greatest number of users are able to pitch in without creating a mess of competing answers.
  • It has been contributed to (and heavily reorganized) by community moderators and some of the highest-rep users on SO, all of whom have greatly expanded moderation tools at their disposal.
  • Despite the continued effort to maintain that particular question, it was created a long time ago. This was a very different site in 2008 and much of what was good practice then is now effectively verboten.

I still think it's a good idea to create this kind of resource in our tag wikis. That being said, it's a massive undertaking for a feature that can support a healthy site but which is not a substitute for core content (questions and answers). We have only three non-moderator users who can approve tag wiki edits and our main challenge right now is to grow our community and expand our core content. So in my opinion, it would be best to table any major effort for now, and revisit this idea when the beta picks up steam.

$\endgroup$
  • $\begingroup$ "[it'll] give you an idea of whether this community is ready to take on the effort you're suggesting". The community has already spoken, and the answer is nope. $\endgroup$ – user3791372 Aug 21 '14 at 19:11
  • $\begingroup$ @user3791372 Less than 24 hours is not sufficient time to reach consensus, and the verb here is would - as in, it would give you such an idea, if you suggested this particular alternative to your current approach. $\endgroup$ – Air Aug 21 '14 at 19:16
  • $\begingroup$ @user3791372 I found an example of this type of resource on SO and added to my answer. For what it's worth, if you're interested in making a real, comprehensive effort to spearhead this, I think you can absolutely earn the support of the community (and mine). But I think you could make a bigger impact at this time in other areas. $\endgroup$ – Air Aug 22 '14 at 17:04
  • $\begingroup$ That C++ book thread was the exact thread I was trying to find after you posted this answer, it gave me the idea. Such a question and resource would most certainly pull in lots of additional people. Still, the reversed decisions after the eating of humble pie doesn't change first impressions and I think the down votes speak for forseeable community support. Writing of which, I notice about 7 down votes (from my meta question and this question) have just suddenly vanished. $\endgroup$ – user3791372 Aug 22 '14 at 17:28
-1
$\begingroup$

Link-only answers are discouraged, and so are the questions that raise such answers. Thus, as long as your question asks not only for links, but also for minimal descriptions of the courses, be it about the quality/content/duration/..., such list can be very useful for data science beginners.

Considering the purpose of the question, it seems very legit to me. Yet, seeing your post/answers, I think you should make the goal more explicit, or else such undesirable answers will be the only ones coming.

$\endgroup$
  • $\begingroup$ Yeah, guess I was thinking of adding links with description grouped by topic. I think that's why it ought to be a community-wiki answer $\endgroup$ – user3791372 Aug 21 '14 at 11:36

You must log in to answer this question.

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