To identify pertinent theoretical criteria for organizing your variables, read the published literature about your topic, paying attention both to theoretical frameworks and to empirically established patterns. For example, the literature on health services utilization classifies independent variables into predisposing, enabling, and need factors.

If previous articles fail to organize items well or you are studying a new area or using new measures, empirical sorting might help distinguish clusters of variables with similar response patterns. You can then evaluate those groupings to see whether they align with useful conceptual distinctions, as occurred with the abortion items from the General Social Survey.