wow, can't believe China Study is still being referenced. It's been a while since my data stats class, but it was used as a case study for how data can be manipulated to explain a pre-conceived idea.

From memory entire counties were removed that didn't fit, one I recall ate more animal protein than whats found in Western Diets, yet had lowest incidence of heart disease, and some statistical correlations of higher significance were removed, because they didn't fit the narrative.