http://media.noetic.org/uploads/files/Effects_of_choc.pdf Scientists take chocolate, which is then blessed in various ways (by the mediators themselves, a Tibeten shaman, or an intention recording and playback device). The chocolate is then eaten for a week by four groups of 15 people, and their degree of mood disturbance is recorded every day. The results are pretty much the random walk you would expect from such small sample sizes. On day five, some people in the control group were evidently having a bad day. The researchers take this as a significant result. The "P-value" is basically the probability that the results might have been the result of chance, if all other variables are eliminated. The researchers ignore the facts that the groups were actually very different (inevitably, as there were only 15 people in each group), and experienced very different weeks. If you were to take 60 people in San Francisco at random, and separated them out into four groups of 15, you'd likely get very different groups. And a group of 15 (the control) is likely to vary more than a group of 45 (the average of the "intention" groups). So, yes, it is actually easily attributable to chance. What really struck me though, was the extraordinary "get out of criticism free" clause at the end of the paper. So in order for the experiment to work: All partipants must believe it works Everyone aware of the experiment must believe it works (even if they don't participate, or are even nearby) You must tell nobody skeptical about the experiment in the future, or their skepticism will travel back in time, and negate the results. Basically, every time the experiment does not work - then it's because someone was skeptical, somewhere in the world, and possibly hundreds of years in the future.