Herbert Blumer asserts that concept formation is of paramount importance for arriving at valid research results that reflect the reality under study. However, the relationship between the concept and the empirical social reality is a common methodological problem in social research. In this paper, presented in August 1953 at the annual meeting of the American Sociological Association, Blumer addresses this relationship between theories (concepts) and data (the empirical reality), and explains the problem of the "ambiguous nature of concepts" in social theory, and offers suggestions for overcoming it. Among these suggestions is the reliance on sensitizing rather than definitive concepts, since sensitizing concepts gain their usefulness and importance from typical relationships rather than from quantifiable associations.