Overreliance on statistics, and even faith in them--which Richard Smith in this essay calls a branch of "metricophilia"--is a common feature of research in education and in the social sciences more generally. Of course accurate statistics are important, but they often constitute essentially a powerful form of rhetoric. For purposes of analysis and understanding, they have their limitations. In particular they tend to tell us more about correlation than causality. The extended example Smith discusses here--"The Spirit Level: Why More Equal Societies Almost Always Do Better", by Richard Wilkinson and Kate Pickett--shows that while statistics can be valuable guides to where further investigation needs to take place, such investigation needs the tools of other disciplines, such as philosophy and anthropology, if it is to discover matters of fundamental importance.