3 ms·

Are you sure that traditional sources of knowledge, and traditional ways of making people believe things, result in greater accuracy? Consider that over half of

Are you sure that traditional sources of knowledge, and traditional ways of making people believe things, result in greater accuracy? Consider that over half of the papers in many fields, including medicine, are statistically bogus: http://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.0020124 http://journals.plos.org/plosmedicine/article?id=10.1371/jou... http://biorxiv.org/content/early/2016/08/25/071530 http://biorxiv.org/content/early/2016/08/25/071530 On average the results in science papers are probably more accurate than random stuff you read on the internet. But I suspect the difference in error rate is uncomfortably small, much smaller than those who are smug about science would like to believe.

kvb 10y ago
While it may be true that 50% of papers are wrong in some fields, the papers that are wrong are probably not the most highly publicized ones or the ones in the most prestigious journals (and when those papers are wrong the incentive to publish corrective papers is high), so this probably isn't quite as big a problem as it's made out to be.
- laretluval 10y ago
  > the papers that are wrong are probably not the most highly publicized ones or the ones in the most prestigious journals If you google around about the replication crisis you will discover that the opposite is true.
  - fiddlerwoaroof 10y ago
    This is a fairly good example of the sorts of biases one even finds among scientists: http://www.npr.org/sections/health-shots/2015/01/12/375663920/the-doctor-who-championed-hand-washing-and-saved-women-s-lives http://www.npr.org/sections/health-shots/2015/01/12/37566392...
lordnacho 10y ago
The kinds of things that are in current journals are on the cutting edge. We should expect a fair few lines of inquiry to turn out to be negative or wrongly done. The sort of knowledge I'm talking about is the kind of stuff that's been known for decades, acknowledged by everyone who's studied it.

kvb 10y ago

While it may be true that 50% of papers are wrong in some fields, the papers that are wrong are probably not the most highly publicized ones or the ones in the most prestigious journals (and when those papers are wrong the incentive to publish corrective papers is high), so this probably isn't quite as big a problem as it's made out to be.

laretluval 10y ago
> the papers that are wrong are probably not the most highly publicized ones or the ones in the most prestigious journals If you google around about the replication crisis you will discover that the opposite is true.
- fiddlerwoaroof 10y ago
  This is a fairly good example of the sorts of biases one even finds among scientists: http://www.npr.org/sections/health-shots/2015/01/12/375663920/the-doctor-who-championed-hand-washing-and-saved-women-s-lives http://www.npr.org/sections/health-shots/2015/01/12/37566392...

lordnacho 10y ago

The kinds of things that are in current journals are on the cutting edge. We should expect a fair few lines of inquiry to turn out to be negative or wrongly done. The sort of knowledge I'm talking about is the kind of stuff that's been known for decades, acknowledged by everyone who's studied it.