Wednesday, December 21, 2011

Guest post: A curious case of cherry-picking data for the greater good

Guest post: A curious case of cherry-picking data for the greater good

On the subject of Fukushima hysteria, here is a guest post by Alexey Goldin. It's a rebuttal of claims, by activists Joseph Mangano and Janette Sherman, that lethal amounts of fallout have reached North America, and that there is a "dramatic increase" in infant mortality in US cities. (!) Of course they have no evidence for a causal link, just correlation. Of course they didn't explain how on earth the microscopic amount of fallout observed in the US could cause any health effects, let alone lethal acute radiation poisoning; or how the obvious symptoms of such a deadly disease are going completely unnoticed. They haven't merely not-demonstrated causation -- they haven't even hit plausibility.
In this vein I think it is overgenerous to consider the numbers behind these epidemiological claims at all; what would a statistically significant correlation mean, if anything? Extraordinary claims demand extraordinary evidence. Nevertheless, Alexey takes on the epidemiological correlation at face value, and finds it substanceless. According to his analysis there is no statistical significance to the alleged "dramatic increase in mortality". It is noise.

No comments:

Post a Comment