Richard Hughes

This is a wonderful explanation of modern biostatistical techniques for causal inference and bias reduction in the analysis of observational data. One of its great strengths is the consistent use of a mathematical notation. In contrast to reading papers by Rubin, Pearl, Glymour and others where I have to switch from one notation to another, this book presents many importance concepts using a single notation. This makes the book much more understandable. Even better, the description of IPTW is the first time I actually understood IPTW and the logic of why it works. I have read the original MSM paper many times and never understood the true logic of IPTW until I read this book.