A Vegan No More
Great stuff in there. E.g., from the first one:
"Suppose we were to do a study looking at cancer rates among vegans and we find that in our sample we looked over a hundred different types of cancers, and find that for six of them vegans had significantly lower incidences of those cancers at a 95% confidence level. We cannot conclude from this evidence that vegans have lower incidences of those six cancers. This is an example of the Texas sharpshooter fallacy. We have looked at a mass of scattered evidence and then simply picked out those which happened to look like a pattern in our favor."