Network Analysis of Workforce Development Programs

Network structure is one explanatory variable to explain network performance, but its effect may be uncertain. Much research, however, has not attempted to combine network structures with other characteristics of networks – e.g., network social capital, network management, or network learning – that may operate to explain network performance. The consideration of other network characteristics as well as the combination of network structure with other factors cannot be ignored. This research asks the following questions: What characteristics of a network influence its performance? What relationships exist between network characteristics and network performance? and, What is the extent of influence that each network characteristic has on network performance? This study proposed four characteristics of network and presented an integrated model of network performance. The model included unique aspects (structure, content, process, and management) of networks from four network perspectives. This study first examined simple relationships between each characteristic and network performance, and then extended intermediated effects of each characteristic on others. Based on this approach, 25 research hypotheses were proposed. The research site for this study is the workforce development network in Florida. Three groups (rural, suburban, urban areas) are chosen in terms of similar socio-economic conditions. Data are drawn from surveys and interviews with service providers and employers with respect to frequency of contacts (network structure), trust and legitimacy (social capital), adaptation process (network learning), and stability (network management). This study employs a structural equation model after calculating factor loading scores for each dimension. Data of network structure are analyzed with a social network analysis program (UCINET 6.0) and the rest of data is analyzed with the LISREL 8.0 program. Fifteen of the twenty-five hypotheses proposed in this study were supported. First, this study suggested that more diverse service networks or less centralized networks are more likely to lead to better network performance. Second, there is a positive relationship between network social capital and network performance. This finding is consistent with