Identifying opinion leaders for an opinion leadership intervention is critical in order to determine whether or not the diffusion of innovations succeeds; however, there have been few studies that provided a synthetic guideline for researchers to consult in identifying opinion leaders. Based on this necessity, the purpose of this dissertation was to provide a synthesis of the ways to identify opinion leaders by using methods of social network analysis. In particular, this study synthesized communication network data collection methods to identify opinion leaders by using social network analysis and demonstrated how social network data is practically collected in different social systems.
Based on the purpose of the study and background knowledge about identifying opinion leaders by using social network analysis, this dissertation posed two prime research questions: “What are the characteristics of each opinion leadership data collection method used to identify opinion leaders?” and “How is a social network data constructed by using opinion leadership instruments in opinion leadership research?” Each prime question has three sub-questions.
In terms of synthesizing characteristics of social network data collection methods, this dissertation categorized five methods–sociometric, informants’ rating, snowball, observation, and self-designating methods–through a literature review of opinion leadership studies. In addition, this dissertation presented two empirical studies to demonstrate the procedure of how social network data could be constructed as the second prime research question asked. The first empirical study was to identify opinion leaders within Pennsylvania’s juvenile justice system and the other was to identify opinion leaders within fourteen Indian agricultural villages. Both studies were formative studies for purposive diffusion projects. Related to further methodological development of social network analysis to identify opinion leaders within a social system, some suggestions were offered.