NewScientist: Most scientific papers are probably wrong

[Telos]

NewScientist: "Most scientific papers are probably wrong"

http://www.newscientist.com/article.ns?id=dn7915

Most published scientific research papers are wrong, according to a new analysis. Assuming that the new paper is itself correct, problems with experimental and statistical methods mean that there is less than a 50% chance that the results of any randomly chosen scientific paper are true.

This is no surprise to me if we are considering papers in the social sciences. But I wonder what those of you in the other sciences think about this?

 

[Ivan Seeking]

Somewhere out there, there is a paper that uses figures on the strength of a certain dental material that were obtained by me in my garage. I find that to be rather scary!

 

[Pengwuino]

Somewhere out there, there is a paper that uses figures on the strength of a certain dental material that were obtained by me in my garage. I find that to be rather scary!

And I built my garage with that dental material!

 

[Ivan Seeking]

And I built my garage with that dental material!

As long as you don't bite it you'll be fine.

 

[Pengwuino]

mmm... that might be hard. I didn't have a lot of plaster... so i used cookie dough.

 

[LeonhardEuler]

http://www.newscientist.com/article.ns?id=dn7915

Most published scientific research papers are wrong, according to a new analysis. Assuming that the new paper is itself correct, problems with experimental and statistical methods mean that there is less than a 50% chance that the results of any randomly chosen scientific paper are true.

This is no surprise to me if we are considering papers in the social sciences. But I wonder what those of you in the other sciences think about this?

This result was obtained in a scientific research paper. The results of scientific research papers have a greater than 50% chance of being wrong. Therefore, there is a greater than 50% chance that the conclusion is wrong. This implies that there is actually not a greater than 50% chance that it is wrong. Since it is probably not wrong, then the conslusion is probably true, which means it is probably wrong...

 

[Telos]

Yeah, emphasis on the "assuming that the new paper is itself correct" in the original article.