Comments for Why Most Published Research Findings Are False

Trackback in article The Dark Side of science from N=1: Population of One

2006-01-30T14:15:50Z

While Jack Vinson summarises an article on the unreliabili... [Read More]

Comment from Jack Dahlgren on 2006-01-31

2006-01-31T19:08:52Z

Jack,

Of the six "commonly-held beliefs" that you list, I can't imagine any of them being commonly held. Small sample size has always been a red flag. Financial interest has always been a red flag. Flexibility in outcome means you redefine your result in terms of the data that you have rather than proving or disproving they hypothesis that your experiment is designed to test (presumably your experiment is not as well crafted to prove or disprove some unforeseen hypothesis). Basically all of this is what I'd consider bad science and I'm not even a professional scientist.

It troubles me to see that such things are apparently so common. But it is not surprising, only sad.

Comment from Bill Brantley on 2006-01-31

2006-01-31T23:23:27Z

Check out the latest issue of Harvard Business Review. They have a top ten list of breakthrough business ideas for 2006 with number six being "science in the wild." This fits in well with the article you mention because much of scientific research is done in carefully controlled artificial environments called laboratories. According to the article, complex phenomena cannot be really studied by isolating them from the interactions of other complex phenomena in their natural environment.