You've seen it all over the news, a study tells us that published studies funded by big business are more likely to find something.
OK. You know, if you look at published studies funded by the government they are also likely to find something.
Is this a big conspiracy? Well, no. In case you didn't know, studies that find nothing are very unlikely to be written or published. (Big epidemiological studies that find nothing interesting at all frequently are published, for whatever reason.)
This is a well-known phenomenon in science called "the file-drawer" problem. That means that the data form studies that showed no effect typically languish in a scientist's files. We (scientists) all have drawers full of completely uninteresting, unpublishable data.
Why don't we publish these studies? Well, it's typically because even if we have a sound hypothesis and a rigorous study design, sometimes studies don't yield good data, and if there's no effect we can't always state with total confidence that the "no effect" may not be due to something other than no effect (yeah that gives me a headache too).
To the best of my knowledge, most academic institutions require as part of the contracts with industry that the data are publishable regardless of the final result of the experiment. I myself have worked with industry funding (as well as NIH funding and military funding) and NOT ONCE have there been any restrictions on my funding or any influences on my experimental design based on study outcome. Furthermore, since most of my data come from computer-based scoring and much of my research is conducted by "minions" (or research assistants if you want to be technical) who don't know what the outcome should be, it's nearly impossible to bias that data. Heck, I get paid no matter what the outcome of the experiment is, and trust me--my file drawer has some pretty uninteresting data in it.
You'll notice that Dr. Ludwig and colleagues limited their review of the literature to the health effects of milk, juice and soft drinks. Hmm. They are also researchers interested in preventing childhood obesity. Hmmmmm. On the surface, this could look like bias. Would they be touting this alleged bias if they had included the apple, blueberry, or mineral water industries? Is the discovery that there are beneficial polyphenols in pomegranites "good" if paid for by the NIH and "bad" if paid for by POM?
Are there unethical scientists? Yes. In every field, and they are recipients of funding from all different sources. They are also very, very much a minority.
kOvT1q
-- Posted by: Gmlwatew at July 15, 2009 7:59 PM