"a peer-reviewed analysis by ACSH, using well-established principles of scientific investigation, shows that only one of the 42 sites meets the scientific definition of a disease cluster...
- "All but five of the purported clusters lack any supportive evidence at all;
- "13 examples cited by NRDC as confirmed disease clusters are in fact places where investigations by public health officials are still ongoing;
- "Several cases have already been ruled out by public health officials as disease clusters as no statistically significant elevations of any disease have been found;
- "Only four of the NRDC-claimed sites are possible clusters, and just one of the 42 can actually be identified as a disease cluster based on the scientific definition."
This kind of thinking can be difficult - it often doesn't follow intuitive human reasoning tendencies, especially conclusions based on statistically-based evidence. But good education in reasoning and especially reasoning about statistics would make a world of difference.
Someone who's been thinking a lot about how to improve students' "statistical literacy" is Professor Milo Schield at Augsburg College. Here he has a wealth of information and resources about this issue. I suspect the people at NRDC need a course from him - in the hope that they weren't bending the facts on purpose.