Patterns of sensitivity of association measures in numerical taxonomy (Q790734)

I report here a general model, for a class of measures, that reveals basic patterns of sensitivity that underlie a wide variety of common measures and shows that there is a continuum of potential association measures which exhibit useful combinations of sensitivities. I distinguish between ''separation'' sensitive and ''minimum value'' sensitive measures and describe a new measure that is intermediate in exhibiting both minimum value and separation sensitivity. The utility of this new type of measure in three disciplines - ecology, psychology, and systematics - is briefly described.