I came here because of that. I suspect it's mislabeled, since it looks like it would match the numbers if the labels at the bottom went low, mid, high instead o

I came here because of that. I suspect it's mislabeled, since it looks like it would match the numbers if the labels at the bottom went low, mid, high instead of low, high, mid. But even more than that, I think the numbers demonstrate that some fluoride is better than none. And the low groups average is likely skewed by a single outlier, way at the top of the graph. Poor study all around.