| Chapter 2: | Theory |
The usefulness of network types for analytic purposes has been recognized only recently: “The network type variable is parsimonious and comprehensive. It reflects the social world of each respondent and the varying extent of social support that is perceived to be available” (Litwin, 1998, p. 600). By employing cluster analysis, our procedure constructs four distinct network types (discussed in detail in chapter 5).
Network Content
The second construct in the model refers to the content6 of exchanges between the focal person and members of the network. The exchanges involve the flow of social support within the network and how they are affected by network structure7. As yet, there has been little agreement on the array of exchange items that best reflects the nature of support available; as a result, studies vary from 1 or 2 items to over 50. Our procedure takes a somewhat middle course in terms of the number of items elicited from respondents. For this purpose, we developed the inventory of network transactions (INT) which encompasses some 20 exchange elements. We divide these into three major domains: material, service, and emotional (see Table 2.1). Generally, in the literature, the material and service domains are referred to as instrumental support, and the emotional domain as expressive support (N. Lin, 1986; Wenger, 1997). The transactions between individuals within a network are examined by means of all the exchange modalities potentially available to the focal person. Like the network type construct, the INT puts the emphasis not on individual attributes, but on relations within the immediate interpersonal environment in which individuals are embedded.
While a major focus of the analysis is on the effect of network types, we also take into account any differences between the two national samples (not shown in the model). This allows two basic comparisons throughout the analysis—one based on the distinct cultural background of the two country samples, the other based on the structural differences inherent in the network types. By framing the problem in terms of structural differences (i.e., network types), as well as cultural differences (in this case, between Chinese and American culture), an attempt is made to embrace two often conflicting impulses in the anthropological literature on aging: on the one hand, the comparative method which demands theory building on a universal basis and, on the other hand, the holistic approach which gives priority to understanding aging within its specific and unique cultural context (Fry, 1999).