How do we know where to draw the line at skeptics.SE? In other words, what tests can we use to know we are 1) being skeptical enough, as in sufficiently skeptical (a "not too cold, not too hot, but just right"), and 2) helping encourage free-thinking (avoiding confirmation bias, or perhaps more importantly even community bias, given the community Q&A model)?

Theoretical Example #1: a Q&A is completed, but something is lacking in the accepted answer, leaving room for further skepticism (it answers the question, but before much more thought leads to other questions or re-questioning the original question).

Theoretical Example #2: based on the available evidence/sources, it is clear there are two (or more, e.g. multiple) and distinct schools of thought operating on the same body evidence. Hence we now have the possibility for two valid answers citing the same evidence. How can we support identifying the (or a) better answer?

In a nutshell: How do we help users raise the community level of critical thinking? Can we document here a bullet-point list of tests to help ensure quality of answers (although I'm sure some might help with questions, too)?

List of possibly related questions (esp. the first few):

  • I want to make sure this question is a good one and conveyed correctly, so please feel free to comment with questions or suggested edits if something doesn't make sense/sound quite right.
    – tniles
    Commented Jun 1, 2016 at 0:18
  • What do you mean by "skepticism"? Each source of evidence should be evaluated based on its quality (e.g. a preliminary finding doesn't really "prove" any claim, at best it "hints" that a claim is valid). Voting takes care of both the quality of evidence but, most importantly, of the quality of what the OP takes from it. In other words: by looking at evidence we are practicing skepticism. By "voting", we are increasing the signal-to-noise ratio. Skepticism is not meant in the sense of "trust no one", but "trust the evidence".
    – Sklivvz
    Commented Jun 1, 2016 at 0:54
  • @Sklivvz that makes sense (I appreciate the SNR idea!). Yes, the "trust no one" gets no one anywhere (and fast). Can you address confirmation bias? For example, if a question (or a Q that allows a A which) is either so localized/specific (pigeonholed?) that it only caters to, say, "one side of the aisle" in a political context. Sure, you'll get an answer, maybe two (one from a winger and one from a moderate, both seeming/sounding acceptable)--but at least one will have confirmation bias all over it either way. How would "trust the evidence" compensate for something like this?
    – tniles
    Commented Jun 1, 2016 at 14:30
  • All I can hear is "Can we post a creationist point of view next to a scientific point of view, and regard both as equally right?", in way, way too many words... Commented Jun 12, 2016 at 4:36
  • @gnometorule Not helpful, nor constructive.
    – tniles
    Commented Jun 15, 2016 at 18:59

1 Answer 1


A good test is: is the conclusion based on the best evidence and sound logic.

  • Dawn, thanks. How would a typical user of this site be able to identify and evaluate "best"? E.g., when there are different flavors of conclusions, as is the case if you read all the A's in a Q with multiple A's listed. It just seems like a possible (highly likely?) area for subjectivity to creep in if most are casual readers, not knowing how to compare evidence/sources. Maybe this sounds like a nuance, but I don't think so. I hope everyone agrees with sound logic as a companion to evidence. Expounding on that may be outside the scope of this question, but it may also help future readers.
    – tniles
    Commented Jun 1, 2016 at 14:39
  • meta.skeptics.stackexchange.com/a/2528/30557
    – user30557
    Commented Jun 1, 2016 at 14:43

You must log in to answer this question.

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