In reference 6 from the Wikipedia article on Radiation Hormesis, the abstract says "Data concerning about 85,000 exposed animals and their 45,000 corresponding

In reference 6 from the Wikipedia article on Radiation Hormesis, the abstract says "Data concerning about 85,000 exposed animals and their 45,000 corresponding controls, with a total of over 60,000 and 12,000 cancers in exposed and control animals, respectively, have been collected." So, for suitable definitions of controlled environment, they have been repeated. I'm frankly tempted to believe this data. There does seem to be a plausible mechanism of action, and there isn't, to my knowledge, much evidence to prefer the Linear No Threshold model. On the other hand, I probably won't go and tape a radioactive source over my pancreas, so I don't believe either model enough to actively do anything about it.