Abstract
Research problem: For knowledge workers, such as IT professionals, the ties within their social network are a major channel for communicating the requisite knowledge and information. While prior studies tended to favor higher network centrality (i.e., larger volume of network ties), in this study, the researchers argue that centrality can reduce communication efficiency if the diversity of the ties is low. Research question: Which characteristic of communication ties, volume or diversity, has more influence on knowledge workers' performance? Literature review: Using social network analysis (SNA) as the theoretical framework, a review of the literature shows that social network ties have important performance implications because they can enhance an individual's access to valued resources. Studies have also examined the performance impact of “hindrance network centrality,” or how frequently a person is described by other network members as a hindrance to their performance. However, current research has overlooked the possible negative communication consequences of centrality in regards to redundant information, which may negatively affect job performance. Methodology: The current study employs a quantitative approach, using the standard SNA method of a “name generator” questionnaire to collect network data. Participants were 98 people in the IT department at a large defense company. We identified 3905 relationships and analyzed them with the UCINET software. Subsequently, social network variables, as well as performance and human capital variables were analyzed in logistic regressions. Results and discussion: Results showed a significant relationship between constraint, a measure of tie diversity, and performance. Centrality, the measure of tie volume, however, was not significantly related to performance. Our findings about the volume versus diversity of communication ties have important implications for professional communicators in the age of fast-growing social network media. A limitation of our study is that we approximated communication ties with social network links. We suggest that future studies further validate our findings by using a more direct measure of communication ties.