This article evaluates a number of different methods for summarizing the relationship between prevalence and other variables. It uses as a concrete example the article by Jorm et al. The article makes recommendations for future studies.