granularity

One of the judgments made by scientists is how to aggregate or segregate data – especially when it comes to changing a continuous variable like age – into separate bins (10 to 18, 19 to 34, etc.). Race/ethnicity as a category has come in for some well-deserved criticism. Leave aside the argument that it is a social construct, race/ethnicity contains too many confounding features. A study in Nature points to a new way to break the category into meaningful segments.