11
$\begingroup$

Earth Science Stack Exchange is a science site, and, in principle, we require or strongly encourage sources for answers.

It's been nearly six years since the linked meta post, which was in fact posted on the first day of private beta. We're a different community now and I would like the (renewed) community opinion on how strict we should be in enforcing the scientific nature of our site.

How strict should we be in enforcing this?

| |
$\endgroup$
  • 1
    $\begingroup$ I cannot vote for any of gerri's proposed policies. Two are far too harsh, the third a bit too soft. What I'd like is a "Goldilocks zone" policy. $\endgroup$ – David Hammen Feb 3 at 13:00
  • 1
    $\begingroup$ @DavidHammen You're very welcome to formulate a goldilocks zone answer, my initial three answers were by no means intended to be comprehensive, but rather just a start. $\endgroup$ – gerrit Feb 3 at 14:07
  • 1
    $\begingroup$ There is a tension between wanting people to provide answers and wanting them to provide well documented answers. I am less likely to provide information if it requires me to do research to make it complete. In some cases it is easy, but in some cases it is not. I think an answer is valuable if it provides enough that it can help the person who asked the question do their own research like, "Research on such and such location has shown it has the following properties which answers your question." $\endgroup$ – haresfur Feb 14 at 2:24
11
$\begingroup$

I find nice the message the system gives when citation is needed.

enter image description here

I particularly don't think the notice should only apply to downvoted question.

This site is plentifull of scientifists users. It is common sense. If I say limestones are sedimentary rocks no citation is needed, I am not going to quote Charles Lyell. But if the statements are related with recent studies, those studies normally are quoted.

What I wonder is when moderators should notice it. I think we users can help on comments. eg "can you provide sources of your afirmation x?". If the poster don't provide sources, then moderators could have it clear: some users asked for sources; poster didn't provide sources.

We add the notice as we don't really know if the information is true.


gerrit a moderator here notize you can flag the post and ask for this "need sources" pop-up, so moderators review it.

| |
$\endgroup$
  • 5
    $\begingroup$ Note that anyone can flag any post and request for this notice (or any other notice) to be added. $\endgroup$ – gerrit Jan 31 at 13:56
4
$\begingroup$

We should be moderately strict. Any answer without citations will receive a post notice. If not improved, it will ultimately be removed.

| |
$\endgroup$
  • 2
    $\begingroup$ I voted for this because some answers are by nature subjective and thus impossible to back with references, like answers to outcrop interpretation questions for example. $\endgroup$ – Jean-Marie Prival Jan 29 at 14:52
  • 4
    $\begingroup$ I would generally accept this (I voted for both versions), but some answers may be really basal, and I must admit, on one or two occasions I interwove an assessment that I didn't back up with a reference. $\endgroup$ – user18607 Jan 29 at 20:12
  • 2
    $\begingroup$ "ultimately removed" is too aggressive. The post notice itself serves to warn future readers that there is some question. If there is some factual problem with the answer (e.g. it's wrong) then it should be removed, but if the answer is not disputed at all and only lacks a source, the notice would probably be sufficient. $\endgroup$ – uhoh Jan 30 at 0:01
  • 3
    $\begingroup$ @uhoh A bit like (English language) Wikipedia's noticed "citation needed" then? $\endgroup$ – gerrit Jan 30 at 8:46
  • 3
    $\begingroup$ I think we should recognise that there's a difference between an answer which doesn't cite references because it is basic information, and one which makes sweeping statements without support. There are a number of users who routinely provide poor quality answers with no supporting evidence, and still manage to gain rankings, because they are the first to answer, or the OP is just being polite. There should be some form of penalty for this as it reduces the overall quality of answers on the site and provides a perverse incentive to keep posting crap. $\endgroup$ – Will Jan 30 at 21:06
  • $\begingroup$ @Will One upvote is worth five downvotes in reputation... that's not something we can change here. $\endgroup$ – gerrit Jan 30 at 22:08
  • 3
    $\begingroup$ @gerrit I understand that, but the post notice does indicate that an answer may be problematic - e.g. no citations, and fwiw I think we should be using that more. In the case of persistent offenders, I'd consider deleting answers, because they are not following best practice, and the site is unlikely to grow in the way we want it to if we don't work to improve the quality of content. $\endgroup$ – Will Jan 31 at 8:43
  • 1
    $\begingroup$ I voted to this. It would be nice a tip type Wikipedia: "citation needed", but I agree there are some questions that do not need references, so: <s>Any answer</s> Some answers without citations will receive a post notice. But this would give extra work for mods, and they would need to decide where the post needs citation or no. We should have some kind of rules about when does a post need citation. $\endgroup$ – user18590 Jan 31 at 9:21
  • 4
    $\begingroup$ I have started adding some "citation needed" posts to some recently answers with poor votes. $\endgroup$ – gerrit Jan 31 at 9:24
  • $\begingroup$ This is overly harsh. $\endgroup$ – David Hammen Feb 3 at 13:02
4
$\begingroup$

Responding to Jean-Marie Prival's comment that some answers cannot have references because of them being subjective - even in those answers it is still possible to include references that explain the problem and show where the current state of research is. I have done this in a quite a few of my answers where typically OP asks a world building climate change scenario question. One can also bring in references to scientific blogs and popular science articles.

The real issue here is that some people have found a way to break the SE rating/points system i.e. a way to gain a very high reputation without citing a single reference and I think that needs to be tackled separately.

UPDATE In the answers to this closed question - In which state of matter exactly are the clouds (can only be one of the three stages [closed] I have been told very politely in the comments that references are not needed for this particular question by a couple of people.

May I politely and respectfully disagree. First of all meteorology and atmospheric science is still a young field and sophisticated observations available via remote sensing confirm what we do know. Certainly I do not recall that cirrus clouds were composed of ice when I was introduced to different types of clouds in high school.

High flying aircraft having particle sampling instruments and satellite radars(CloudSat and GPM) have confirmed this so called "textbook" fact.

If I were answering that question I would most definitely try to provide a more modern perspective(depending on OP's background).

TLDR It is not all hand waving and it is not all wikipedia.

| |
$\endgroup$
  • $\begingroup$ Of course it is always possible, but to what extent is it desirable? If I state that granite usually contains quartz and feldspar, do I have to cite Streckeisen and include an image of a QAPF diagram to prove it? $\endgroup$ – Jean-Marie Prival Feb 3 at 9:11
  • 2
    $\begingroup$ @Jean-MariePrival I maybe coming from a atmospheric science perspective. What I usually do before answering a question is to ask OP about their background and then adjust accordingly $\endgroup$ – gansub Feb 3 at 9:45
  • $\begingroup$ Yeah I'll do my duty when able to, since I passed 2000 I spend my 5 daily votes to delete rock id questions... I understand that some privileges need to be earned by proving you're trustworthy, but IMHO there are too many steps in the current system. $\endgroup$ – Jean-Marie Prival Feb 3 at 13:09
  • $\begingroup$ You're right for the background, I have a bias "Earth science = geology", forgetting about the fluid envelopes surrounding the solid Earth... Sorry. $\endgroup$ – Jean-Marie Prival Feb 3 at 13:13
  • 1
    $\begingroup$ Re "some people have found a way to break the SE rating/points system", it would be interesting to know what fraction of SE participants actually care about that. $\endgroup$ – jamesqf Feb 3 at 17:39
  • $\begingroup$ @Jean-Marie-Prival Thanks, myself I started the closing task. I see you are deleting them. From the initial 500 id-my-rock questions there are only 120. I would also delete some Walsby content btw, but my rep don't allows me to vote to delete answers. $\endgroup$ – user18590 Feb 4 at 9:56
  • 1
    $\begingroup$ @jamesqf We solved the issue of id-my-rock and this issue is again giving the feelling our inbox is spammed. The home page is what new users read first. We cannot give the feelling our content is poor if we wish to grow as commnity. He got a suspension for one day. I hoppe he stops writing about everything, about birds, about soils, about trees, about hurricans.... $\endgroup$ – user18590 Feb 4 at 12:17
  • 1
    $\begingroup$ @gansub The same thing had been happening on physics SE; but eventually they seem to have lost interest and left. I had been getting very frustrated by the blatantly wrong things being presented as fact, and then when you question it the appeal is just to "my experience", "common sense" or "something I remember reading". $\endgroup$ – JMac Feb 11 at 15:12
  • $\begingroup$ @JMac they are still active on ESSE and just continuing on and on and on and on. It is a nightmare $\endgroup$ – gansub Feb 24 at 5:37
4
$\begingroup$

We should be moderately lenient. Answers only need citations for controversial claims and/or if requested in the comments, which may trigger the "Want to improve this post?" post notice. The appearance of this notice indicates that the poster is required to add citations to their answer and/or accept edits that add citations on their behalf. Failure to do so will result in the answer being deleted.


My aim with this is to pitch something between "moderately strict", which requires all answers to have citations (too harsh), and "lenient", which currently doesn't indicate the result of non-compliance. I realise that these are all just sketches of what a process would ultimately look like, but it feels like another point on the scale would be useful for voting purposes.

| |
$\endgroup$
3
$\begingroup$

We should be lenient. Answers only need citations for controversial claims and/or if requested in the comments. "Common textbook knowledge" should not normally need a citation and no action will be undertaken if those are missing.

| |
$\endgroup$
  • 2
    $\begingroup$ This. I, and perhaps others, would be deterred from answering things that are textbook level / general knowledge for a subject specialist, if I had to spend time finding a source for them.Additionally, as Jean-Marie Prival mentioned on a different answer, some answers don't lend themselves to citations at all. Where an answer is controversial or otherwise making surprising claims, then it's reasonable to ask for those claims to be backed up. People may be quicker to request that of some answer-writers than others, depending on their history/reputation. $\endgroup$ – Semidiurnal Simon Jan 29 at 16:35
  • 3
    $\begingroup$ Ok. The rule "backup if asked" should be applied to questions, too. I kow it costs time, some answers cost me more than hour +additional search when asked (or criticized). I would find it good if we could suppress the firing off of low quality answers. $\endgroup$ – user18607 Jan 29 at 20:12
  • 2
    $\begingroup$ In addition to the exceptions of 1) controversial claims and 2) if requested in comments, there's also 3) very specialized or specific information where a source should be cited. $\endgroup$ – uhoh Jan 30 at 0:01
  • 1
    $\begingroup$ I prefer it when answers do use citations, but it's not always possible, for all the reasons mentioned so far. Sometimes a question is asked that is very esoteric & there are limited references, if any, particularly on line to use. The question, Ore Reserve Definition is an example. I answered that based on accepted work practices & what I learnt years ago as a student. $\endgroup$ – Fred Jan 30 at 9:45
  • 6
    $\begingroup$ What irritates me is when someone, who is prolific in answering questions, never includes a citation in any answer. When challenged the individual states they don't know how to include citations & it is apparent the person doesn't want to learn how to. The person only posts answers & questions with nothing but text. I think we need to be a bit more insistent with such people & if their answers do not have citations when requested, the answer should be deleted. $\endgroup$ – Fred Jan 30 at 9:54
  • 6
    $\begingroup$ I agree with @Fred there's one poster in particular who routinely posts rambling answers with no supporting evidence - with additional off-topic snide comments. Allowing this kind of behaviour reduces the quality of answers on the site, and risks discouraging new, more complete, or factually accurate answers. It may be that we need some form of quality control beyond a simple downvote (particularly as the newest users don't have this privilege) $\endgroup$ – Will Jan 30 at 21:11
  • 2
    $\begingroup$ @Will while I share your frustration with this individual, down-voting does work: the answers tend to have negative scores and be greyed out. Sadly the individual keeps accumulating rep because upvotes are worth more than downvotes, but nothing here would change that. I agree about being more liberal with the "citations needed" box - I think only mods can add that? $\endgroup$ – Semidiurnal Simon Jan 31 at 12:53
  • $\begingroup$ I think we can flag it to add one? $\endgroup$ – Will Jan 31 at 19:36
  • 1
    $\begingroup$ @Will Yes. Please do so if you think it is needed. $\endgroup$ – gerrit Jan 31 at 20:53
  • 3
    $\begingroup$ @gerrit - The current choices when flagging are "spam", "rude or abusive", "not an answer", "very low quality", or "in need of moderator intervention". It would be nice if "needs citations" was added as an explicit flag reason. $\endgroup$ – David Hammen Feb 3 at 11:49
  • 1
    $\begingroup$ This is a bit too lenient, but it's close, particularly if the comments are incorporated to formulate policy. As I mentioned above, it would also be helpful if it was possible to directly flag an answer as being in need of references as opposed to flagging it as "in need of moderator intervention" with a comment "citation(s) needed". $\endgroup$ – David Hammen Feb 3 at 13:11
  • 1
    $\begingroup$ The phrase "common textbook knowledge" is a bit fuzzy. Where's the borderline? I would view the text for an earth sciences class kids take as 11 year olds "common textbook knowledge". But what about the text used in an introductory college class? In an upper level undergrad class? In a graduate level class? $\endgroup$ – David Hammen Feb 3 at 13:18
  • 1
    $\begingroup$ It does not help that "common textbook knowledge" can sometimes be very wrong. A 60+ year old text on mountain formation was must likely very wrong; such texts precede plate tectonics. Until very recently, many (most?) college level texts on tides were wrong. Invoking centrifugal force is flat-out wrong, as is the concept of tidal bulges. $\endgroup$ – David Hammen Feb 3 at 13:23
  • $\begingroup$ @DavidHammen Unfortunately, the software is not flexible enough to add extra explicit flag reasons. The flags are the same across the SE network, and the software handles different flags differently. "In need of moderator intervention" is always handled by ♦ moderators, whereas multiple "spam" flags can delete a post even if the flaggers don't normally have the ability to delete, and other flags have special software handling as well. It would be nice if multiple users flagging as "citation needed" would automatically add a "citation needed" post notice, but this is not currently possible. $\endgroup$ – gerrit Feb 3 at 14:12
-1
$\begingroup$

We should be very strict. Any answer without citations will be deleted. This is similar to how Skeptics Stack Exchange works.

| |
$\endgroup$
  • 2
    $\begingroup$ This is far too harsh. I don't participate in skeptics all that often precisely for this reason. $\endgroup$ – David Hammen Feb 3 at 13:03

You must log in to answer this question.

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