for the first point, I think I have seen small/average/large for gestational age being sorted as average/large/small (both on an axis and in legends). For the second it could be something like this (www.flickr.com/photos/65802...)

I agree, the framework is a good (least crappy?) solution for observational studies, but the term TTE itself is like Feathers McGraw's rubber glove. (Some) Reader think they see a friendly chicken when they are in fact staring at a villainous penguin

In the same chapter there is also a 95% CI for a ratio going below 0 - is that a bug or a feature?

Hi, just started reading your book - impressive work. Do you have a place for reporting errors etc. specifically for the book?

I would probably expect that the test below (Chapter 4) was for ratios being equal to 1, but it appears that tests are for ratios being 0. Am I missing a point here?

great post - the final twist reminds me of this classic xkcd (xkcd.com/552/)