Here is an example of what my buddy R calls a BFO (blinding flash of the obvious) via SciAm online (http://www.sciam.com/article.cfm?cha...DD3BFB99CCD7):

__________________
The dose makes the poison--Paracelsus (1493-1541) http://en.wikipedia.org/wiki/Paracelsus

I don't know. That's why I'm asking--Noclevername

Intelligence may not be clearly defined, but you know stupid when you see it--Noclevername

Science is a way of thinking much more than it is a body of knowledge--Carl Sagan (1934-1996)

There is a serious vagueness factor in that. Some studies are paid for just to confirm the obvious.

What would be nice in this is if there was a "level" of complexity stated. It appears as written to be an "average of averages" falicy

__________________
In your rush to call everyone "entrenched" or closed-minded or "limited" you fail to note that the "limit" here has a very natural boundary: that point at which the evidence stops. - JayUtah

Science fiction was never meant to be an educational tool. - Editor Amazing Tales

Well, only 111 of the 206 had data on sponsorship, so about 45% of the sample in the meta-analysis was unaccounted for in their results. That's an enormous percentage! However, 'sponsor bias' in nutritional studies is not a surprising finding, considering the the numbers of recent articles discussing 'sponsor-bias' in clinical trials sponsored by pharmaceutical companies: http://www.drugresearcher.com/news/n...-trials-breast

__________________
The dose makes the poison--Paracelsus (1493-1541) http://en.wikipedia.org/wiki/Paracelsus

I don't know. That's why I'm asking--Noclevername

Intelligence may not be clearly defined, but you know stupid when you see it--Noclevername

Science is a way of thinking much more than it is a body of knowledge--Carl Sagan (1934-1996)

It's funny how people nearly always see leap to the idea that there's a sinister conspiracy such as researchers taking money to lie, but practically never consider or even appear to imagine the idea of an alternative explanation... such as that a company with the truth on its side has an incentive to spend money on getting it out. (It's similar with potlical campaign contributions: they're always described as payments for politicians to obey their contributors' orders, never as people simply sending money to the politicians whose ideas they agree with anyway...)

Nobody is saying there is a conspiracy, Delvo. The possibilities raised in some of the papers I've read on the subject included the possibility that a pharma company would not sponsor a clinical study of a given drug unless that company had a very good idea ahead of time that the drug in question was efficacious. Thus, companies would not sponsor studies on drug candidates that they thought would produce negative results in those sponsored studies, biasing the sample of industry-sponsored studies towards those producing positive results.

__________________
The dose makes the poison--Paracelsus (1493-1541) http://en.wikipedia.org/wiki/Paracelsus

I don't know. That's why I'm asking--Noclevername

Intelligence may not be clearly defined, but you know stupid when you see it--Noclevername

Science is a way of thinking much more than it is a body of knowledge--Carl Sagan (1934-1996)

But I think what BigDon and Delvo are saying is, that's not really a "bias". A bias would be the intimation that the results of the study themselves are somehow contaminated or unreliable, and that is what this data does not demonstrate. Even though you are not saying it does, a casual ear might hear that. Note, for example, the words "unfortunately" and "distort" in the OP quotation-- those words are not supported by the data without better controls. I think what it means is that the industry itself is not asking the tough questions, one needs outside funding for that, but it doesn't mean they are not finding valid answers to the questions they are asking.

Point taken, Ken G! However, I'd call attention to the last part of your post--the kinds of questions one asks, and how they are asked, may unintentionally bias the overall conclusion one would draw from the study, even if the individual questions were answered in a perfectly valid way in the context of the study itself.

That said, this is a problem with many studies regardless of sponsorship, and data from studies examining the relationship between a single nutrient and disease in humans are notoriously fraught with confounders, noisy, and difficult to interpret.

__________________
The dose makes the poison--Paracelsus (1493-1541) http://en.wikipedia.org/wiki/Paracelsus

I don't know. That's why I'm asking--Noclevername

Intelligence may not be clearly defined, but you know stupid when you see it--Noclevername

Science is a way of thinking much more than it is a body of knowledge--Carl Sagan (1934-1996)

As I understand the post by Paracelsus, the bias isn't in the studies themselves, but in the process of choosing what to do the studies on.

__________________
A person's name, or a mark representing it, as signed personally or by deputy, as in subscribing a letter or other document.

I agree, the potential for skewing the interpretation exists just by the questions asked. I think it points to the need not so much to take with a grain of salt what the results are that are funded by the food industry-- the results may be perfectly valid-- but rather to the need for independent funding to ask those tougher questions.

Yes, the studies seem to give contradicting results every ten years. It's as though it takes about ten years for the previous study to seem outdated, so someone asks the question in a somewhat new way, does a new study, and obtains the opposite interpretation. It's a real mess, no doubt.

Part of the problem with reporting of "bias" in scientific studies is not just many people's tendency to jump to conspiratorial conclusions, but also that the word seems to mean something to the general public that it doesn't mean to statisticians or anyone else who uses statistics much. As a statistical term, it just means anything that would (or even might) cause the results to lean in a certain direction. As a "common" word, it's used to indicate more hostile things: deliberate deception or blinding prejudgement.

Yes, that's so-- "bias" has a stigma in common parlance. Another problem is the phrase "sponsorship bias", which probably means "bias in the data that results from the fact that the study is sponsored in a particular way", like choosing to fund studies on foods they feel will have a particular benefit and want to establish that, but it would be easy to read as "bias that is intentionally exerted by the sponsors", like telling the researchers they'll only get more money if the results play out a certain way. Granted, the latter probably does happen (the "duh" element), but this study doesn't establish that.

As one who teaches people to grow food, and think about their choices, I found it more than disconcerting to happen to pick up a paperback book called, "Trust Us, We,re Experts. see:http://www.thirdworldtraveler.com/Public_Relations/Trust_Us_We're_Experts.html

It is clear from the author's work that we are told little of the presence of genetically modified foods in our diet, or that there have been little or no tests on mammals or primates with it, and that when one of the gurus of genetic modification was called in to be an expert witness for a food giant study(I won't name names here)..he couldn't certify it as safe, and his career was carefully disassembled. Disturbing.
As a case in point, the introduction of lectins (a cousin of ricin) into potato tubers from snowdrops (Galanthus) saves agribusiness spraying for Colorado Potato Beetle, but it shows up in every potato you eat. Would you have salt,pepper, and ricin/lectin shakers on your table for guests for a roast beef & mashed potato dinner?
As Paracelcus says...the dose makes the poison. Organic foods are growing rapidly in some areas for good reason. Pete

__________________
A third rate theory forbids.
A second rate theory explains after the fact.
A first rate theory predicts.
A. Lomonosov