Causality

Here's what I don't like about "studies" that conclude something like, "If you eat more than 3 pieces of bread a day, you are 25% more likely to get cancer." These tests always suggest cause-and-effect, causality, but in fact you can compare any two things and come up with some kind of probability based on the data, which does not of itself prove anything more than the relationship of the samples. Thus, I could compare those who have read Homer in Greek and those who have not, in relationship to getting cancer, and say something like, "Those who read Homer in Greek are half as likely to get cancer," and this sounds significant. The misleading factor is grammatical. The results are presented in the present or future tense, but the actual facts merely compare two samples in the past tense. Just because you compare two or more things does not mean there is a causal relationship between them. I bet 90% of these kinds of studies are nonsense in terms of causality.