I see the "graph" on page 104, which shows (if anything) a slightly negative effect for men ("high power" men took slightly fewer cookies on average). I didn't see a detailed methodology anywhere.