LogFAQs > #961429581

adjl posted...

I think you're grossly overestimating the intelligence (or at least the scientific literacy) of the average human. The majority opinion of laypersons means pretty much nothing as far as a given scientific conclusion's validity goes.

That's why I said "you present your data".

The scientists did not, they just went "We believe it because it goes with our research so you should believe us. What's our research? Oh no you're too stupid to get it, so I won't bother to explain, just trust my authority."

Fortunately the majority of people have enough scientific literacy or faith in authority to accept it without all the data given.

adjl posted...

even then its validity is less a matter of how many scientists agree with it (which would be blatant argumentum ad populum) and how robustly the scientific community has failed to refute it. The scientific method boils down to "maybe this is true, let's all try to disprove it and we'll accept it as being a more likely explanation than others if we can't."

Any scientist remotely competent enough to get the qualifications is aware that if they fail to refute it, it is the most accurate answer available at that time, it's bad science to go "I don't have sufficient data to refute the facts, but I choose to disagree despite that."

adjl posted...

Science doesn't actually give you true information. It just gives you information that nobody can prove isn't true. In practice, those are similar enough for laypersons to ignore the distinction, but it's a very important philosophical difference that needs to be considered in critically appraising any given scientific work. Most notably, if a published paper seems to be trying to prove their hypothesis right instead of trying and failing to prove it wrong, that's a glaring methodological problem that indicates there's probably quite a bit of bias in there.

That's what peer review is for, and peer review has shown nobody has a sufficient proof.

The distinction is not meaningful for public acceptance as the goal isn't to be as accurate as possible when you eventually make a conclusion, it's to have an answer at this moment that is as accurate as current data can discern, and change as and when new data appears.