When Foxon and Shiffman confined their analysis to people who could be reasonably described as smokers, they found the opposite of what Glantz claimed.