7

Generally, I feel that most of our questions get very good answers.

We get a lot of answers based upon experiences and anecdotes, and most of them are very high quality. However, I feel that we've been seeing fewer answers that provide references and citations. While I don't want to detract from answers based upon personal experiences and anecdotes, I would like to encourage a more even mix of anecdotal and referenced answers.

EDIT: To clarify, I don't think that answers that rely upon personal experiences and anecdotes are bad, or even inferior to those that provide references. Some questions clearly don't need references or citations, and are, in fact, better answered by personal experience and anecdotes. However, other types of questions (such as those about health and safety issues) may benefit from well-referenced answers. I think our site would benefit from a good mix of both, which is part of my intent here (the other parts being drawing attention to meta, and increasing the number of users we have with higher reputation amounts).

So, in the spirit of encouraging these types of answers, I'm going to be awarding a series of 500 point reputation bounties on question for answers that provide comprehensive, well-researched information.

I will post the details of the bounties I will award, and which question they will apply to, as an answer to this meta question. This way, if there are other questions where I or another user wishes to award similar bounties, the details can be added as an additional answer.

Another issue that has come up has been the lack of attention and activity here on meta. Since you're reading this, you now know about the meta site, and I strongly encourage you to look at some of the other questions and answers here on meta. Please take the time to vote on questions or answers you feel are useful... remember: votes on meta are a bit different than on the main site. Up-votes on questions mean that you think the question is worth discussion. Up-votes on answers mean you feel the suggestion is a good one; downvotes mean you disagree.

Votes on meta don't impact reputation, but they are very important for defining ourselves as a community. Voting here helps our chances of graduating out of beta, almost as much as voting on the main site does!

  • Agreed ! I just arrived here after publishing this answer : parenting.stackexchange.com/questions/1908/… Do you think a source like this one is valuable even if in a foreign language ? I have no idea which amount of our community is actually native english speakers (I'm French). – Michel Daviot Apr 3 '13 at 20:41
  • I think sources are almost always valuable, even if not in English (so long as an English summary of the main points are provided). – user420 Apr 3 '13 at 21:40
4

Being a librarian contributing to a site which has as its mission to "build libraries of high-quality questions and answers," I am all for citations in answers. Some thoughts:

  • There are amazing databases available for free from most public libraries (maybe not all countries). You can log into them for free from home. All it takes is a library card - usually free. Scholar.google.com and books.google.com also have some pretty good resources. There are many government documents and databases freely available (append site:.gov to any google search)
  • Yes, you might find two studies that give differing conclusions, but a good study will review available literature, so the aim should be to find recent studies that have reviewed the studies that came before them.
  • Some questions lend themselves to anecdote (strategies to reduce computer game interest) and others really need authoritative answers (what crib bedding to reduce SIDS).
  • I LOVED the question about toy guns. If it had been opened to anecdote, you would have gotten parents telling tales of sandwiches bitten into the shapes of guns, but just knowing that a few other parents were okay with their kids having guns is not the same as knowing that this has been researched to the point where governments are actually changing their rules in favor of allowing toy gun play from children. This was not the result I expected to find when I went looking, and I was fascinated to learn that this is a real self-esteem/gender issue for boys.
  • I will always try to find something that backs up my opinion, even when an anecdote will do because the information then has a broader base; it is more informational and less opinion.
2

I'm all for references. I admit I lean a little too much toward the personal experience side. However, I'm not sure asking for peer reviewed academic studies is necessarily the best way to go here. First of all, you need a college level of vocabulary and comprehension in order to understand the study, let alone evaluate its credibility.

Second, the most credible of such studies are generally not available online without a subscription, and subscriptions are unlikely to be worth the cost unless you're already involved in academic research, which most parents aren't. Or you can go down to your local university library, but I think people who are apt to do that won't limit themselves to the studies cited here, and won't be slowed much by a lack of cited studies.

Third, in a field like parenting, for every question you can cite two studies that came to exact opposite conclusions. I'm not sure how much value that adds, especially without access to the text to evaluate the studies' methods.

Lastly, the membership of this site doesn't consist of academics. Even offering a healthy bounty, your war toys question only garnered one answer on a site with a 3.8 answer per question average. I think you may have unintentionally turned away some potentially very good, albeit uncited, answers.

I think the site would be better served by encouraging references "more authoritative than blogs, but more accessible than academic studies." These would be sites like WebMD or parenting magazines, secondary sources who commonly cite academic studies themselves, but translate them into layman's terms.

  • The abstracts of many peer reviewed studies are available without subscription, but I take your point. I did state that references do not have to be peer-reviewed studies; reputable blogs and articles are both acceptable, and much more accessible to the public. Perhaps I did discourage some answers, but there is precedent, and it has only been a day, so it is entirely possible there will be more answers. After all, it still has only had less than 50 views. – user420 Mar 28 '13 at 19:58
  • +1, I am joining Karl on the side of personal experience - it's rare that my answers link to hard facts. One big reason is (frankly) laziness; I don't want to spend time digging up sources to support my claims. I know it can be a weakness (most notably when hard facts are asked for!) but I think that personal experience is valuable too. And just like scientific studies, they might or might not apply to the asker's situation. It's not a competition, just a matter of taste. – Torben Gundtofte-Bruun Apr 1 '13 at 20:06
  • I have edited my question to address the valid issues you have both raised. After re-reading my original post, I realized it seemed potentially critical of anecdotal answers, and that was not my intent. I think there's a place for both types of answers on this site, but I also think that certain types of questions are better served by referenced answers (perhaps not necessarily the gun violence test question, though). Similarly, I feel certain types of questions are better served by answers based on personal experience. – user420 Apr 2 '13 at 13:53
1

The first question that will receive bounties for well-researched answers is:

Have studies shown whether toys that feature guns or other weapons prominently promote violent behavior?.

After one week, I will award TWO bounties of 500 points each to the answer that I feel provides the best references and citations for research. Due to limitations on the bounty system, I will award one, then another after, but this will take several days (48 hours after the question is posted before I can even assign the first bounty, and then I believe there is an additional 24 hour minimum before I can assign it).

Once the two bounties are assigned for the best referenced answer, if there is a second answer that meets the requirements, it will receive an additional bounty of 500 reputation.

You must log in to answer this question.