ABSTRACT
Theoretical methods in structural network research require us to specify a model which is then validated by consistent mappings of data onto it. 'Empirical' methods, by contrast, require less a priori knowledge and are most valuable in the exploratory stage of data analysis. This chapter will focus on the latter. We discuss "empirical" methods which are useful in exploring data. The discussion will center around a method, COBLOC, which has been developed in a series of papers (Carrington, Heil and Berkowitz [1980] and Carrington and Heil [1979]). With the presentation of each aspect of COBLOC we will see alternatives which may be used to improve it. Thus COBLOC should be seen as a tentative step toward a new family of empirical methods.
