Lies, Damn Lies, and Statistics

All of you daytime tv watchers out there probably already saw James Frey's recent meltdown to infamy on Oprah. No matter what the Oprah-haters out there have to say, one thing Oprah does have is integrity. She realized her mistake in defending Frey on Larry King, admitted it, and tried to correct it.

The editors of Science are also trying to show some integrity by proposing new authorship guidelines in the wake of Hwang Woo-Suk's miraculous results. The proposed guidelines would force each author listed on a paper to indicate his/her individual contribution to the research and his/her view of the conclusions. For those of you wondering what it looks like inside a scientists brain, the New York Times produced a (probably) fake article illustrating this proposal.

In the spirit of full disclosure, I feel like I should mention that this article is most likely plagiarized from a real journal article by this guy.

This whole faking research thing is nothing new though. In my particular field, Jan Hendrik Schon produced a total of 15 fake papers in the journals Nature and Science in 2000-2001. Now, these journals are supposed to be the premier source of scientific research, but they often contain less Edison and more Barnum. Nature and Science articles are all about big claims and pretty pictures, but the journal editors get easily bored with things like experimental methods and statistical significance of results, which get pushed into supplementary online materials if they are published at all. All of this tends to eliminate reproducibility, the hallmark that distinguishes science from hokum.

Which finally brings me to the point of this entry, I am seeking submissions for my new journal entitled Hokum: Return of the Double McGuffin.

There is something fascinating about science. One gets such wholesome returns of conjecture out of such a trifling investment of fact.

-Mark Twain
Life on the Mississippi