How (Not) to Use Risk Ratios in Sociological Research

A small but growing literature uses the risk ratio as an association or effect measure. Unlike odds ratios, ratios are unaffected by rescaling noncollapsibility bias and straightforward to interpret. However, has one unattractive property that researchers need be aware of: it is not symmetric with respect outcome definition. The between two groups’ probability of success does equal inverse failure. Choosing category a binary use “success” can significantly affect substantive conclusions, particularly in research comparing highly different base rates outcome. authors give examples from discrimination social mobility studies illustrate this point present rules thumb for depending on rate