It is one thing to get caught in a lie, it is completely another when the whole intent was to lie in the first place and then publish how everyone believed you! An obscure 15-person test case panel in Germany yields a badly and INTENTIONALLY flawed report (that went global) that chocolate is good for you: http://io9.com/i-fooled-millions-into-thinking-chocolate-helps-weight-1707251800
This is a VERY GOOD article. Everyone should read it and the Nature magazine story referenced. Even applies to all those forum posts like ' the new iOS update killed my battery life!' Small sample stats can lie and even well trained scientists don't know it. Sent from my iPhone using Tapatalk
@viewfly: Yeah, I've got to admit that I typically get bored very quickly reading anything more than a few paragraphs. But that article was so enlightening. I'm now permanently a card-holding cynic. LOL. (I always sensed the Huffington Post was running loose and furious with their facts like a child with scissors, but all those other supposedly respectable venues was alarming.)
A lot of people are enjoying this article. Thanks KJ! Here is the link to the embedded Nature Article. http://www.nature.com/news/scientific-method-statistical-errors-1.14700 Sent from my iPhone using Tapatalk
As a scientist, with a few statistics classes under my belt, I rarely believe any of these reports. If it seems valid, I'll search out the report and look at the numbers. But virtually any statistic can be skewed to be statistically significant, if you have a small test group and throw out the outliers. Strange - I'm skeptical of research, but give people the benefit of the doubt.
