Conceptualizing heterogeneous categories negatively is sometimes the better choice.

Evidence

Sometimes one has a choice between a conceptualization that is positive and one that is negative. My favorite example is that “qualitative methods” is best defined as not-statistics, exactly because to give a positive list of qualitative methods would be hard given the heterogeneous methodologies that fall under that rubric. — Gary Goertz, Social Science Concepts and Measurement, 471

Categorization

• SHARE: Nerdy dispute-settler for the one methods-instructor friend who argues about what counts as qualitative.