Abstract

Sociology is split into two antagonistic or mutually oblivious wings: quantitative and nonquantitative. Statistics does not occupy a privileged methodological position vis-a-vis qualitative, verbal sociology. Probability is a theory like any other, and each statistical method contains its particular theoretical bias. Such biases should be brought into the open and tested. Statistics may continue to be useful, though, as a substantive theory of change processes in the social world. A reorientation in our views of statistics may bring mathematical and antimathematical branches of sociology back into a common enterprise