11

I have noticed a tendency for responders to positively state opinions as though they were facts.

As an example see this question
Can the sun damage the camera sensor?
Five answers are in the form of unsupported opinions while the sixth answer is based on reported extensive experience that directly contradicts the other five answers.

Answers can generally be categorised as follows:

a) the responder's unsupported opinions
b) reasoned replies that are deductions from well known facts
c) the responder's reported experiences, observations or measurements
d) the responder's assertion that has backing evidence in the form of citations/references.

a) would seem to be the weakest form of reply while d) is the strongest form of reply, c) depends on your trust that the experience was accurately reported and b) is as accurate as the reasoning and the facts they appeal to.

Wikipedia has a policy of encouraging references for all statements of fact.
Should we follow a similar policy?

A counter argument can be made that this site's system of voting and reputation is designed to give weight to unsupported opinions that are likely to be true. In effect it is resorting to argument from authority, the collective wisdom confirms an opinion.

So the question I am inviting you to discuss is this - should we have a policy of encouraging responders to back their assertions with facts, either from experience or from third party references? Or conversely to ask them to prominently label opinions as personal opinions?

See also Good Subjectve - Bad Subjective

1
  • There are newer answers to that old question, some of which actually cite examples (with photos of the damage).
    – Michael C
    Jan 30, 2019 at 21:47

2 Answers 2

Reset to default
6

I'd like to see us take a middle ground approach.

  • I DO NOT want to see every answer that doesn't have citations for every statement responded to asking for cites. This isn't wikipedia.

  • I DO want to encourage better answers than the regurgitated "common knowledge" answers, the "I just know" and the "20 years ago in college my teacher taught me to always do it this way, because he always did it that way when he was in the field for 30 years before he taught, and I've never questioned it." answers.

I think when it's an answer that no one who is "skilled in the art" (to borrow from patent law language) would dispute, then we don't need to bother with asking for cites. But if we get a case like you point out where there's two different camps, then we need to ask BOTH side to back up their statements. Note for example that Jukka makes a well thought out case against glenneroo's logic about P&Ss.

Oh, and the converse... asking folks to label opinion as such... huge +1 on that. It's not hard, it's as simple as starting a sentence with "I think that...." I try to do it whenever I answer with opinion. That should be a given. Basic code of conduct type thing.

8

I don't know if I would make it a policy.

But I sure would encourage people to gently remind posters who don't provide any citations or reasonable evidence for their claims, with a comment.

Photography is technical enough that it doesn't fall into the subjective trap too much IMHO, but if you feel certain topics are, please refer to the Six Guidelines for Great Subjective Questions:

http://blog.stackoverflow.com/2010/09/good-subjective-bad-subjective/

3
  • 3
    This has been my approach. I think this is an area where moderators can and should actively guide the membership towards following SE or site-specific guidelines. Most people eventually get the point, and start quoting their citations, unless they are purposefully trying to be vague.
    – jrista
    Mar 19, 2011 at 1:30
  • 1
    thanks for that reference. I agree with it. It states the problem and the solution succinctly.
    – labnut
    Mar 20, 2011 at 7:33
  • @jr I agree, but just wanted to note that some people get pretty testy when asked to support their opinions: see the recent interchange in the comments at gis.stackexchange.com/questions/7430/…. The more everyone can do to encourage respondents to back up their opinions, the better off we will all be. Maybe downvoting should be used more. (As I have pointed out elsewhere--forgive me if this is repetitive--the difference between an upvote and no vote is 10 points and between up and down is 12 points, so downvoting is a good tool for rating responses.)
    – whuber
    Mar 22, 2011 at 20:32

You must log in to answer this question.

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