In an undergraduate physics course, I was taught that the easiest way to get "the results you want" is to draw the curve then plot the data. That way, you ignore anything that doesn't fit.
CNN recently had a news story about a report by the University of Copenhagen published in Society of Chemical Industry’s (SCI) Journal of the Science of Food and Agriculture
http://www.cnn.com/2008/TECH/science/08/19/organic.cooking.pv/index.html
Basically, carrots, kale, mature peas, apples and potatoes were raised 3 ways in adjacent fields and then fed to animals over 2 years to see about retention of minerals and trace elements. The three were
- organic soil, with manure but no pesticide,
- organic soil with manure and "organic" pesticides,
- similar soil with mineral fertilizers and pesticides.
Result: no measurable difference.
While the researcher was circumspect in her conclusions, her funding agency was not.
So, what's wrong with this picture?
Bottom line: here is yet another simplistic research result that causes distrust of "science".