Using weak ties to understand resource usage behaviors in an online community of educators

We show that weak ties offer a useful theoretical lens for understanding the sharing and usage of community-contributed resources amongst educators in a large urban school district. Community-contributed resources include a rich variety of teaching and learning resources such as lesson plans, presentation slides, animations and simulations. In this research, we consider whether the deduced relationships between members of the community constitute weak ties. A deduced relationship exists when two community members view or access the same resource. If these deduced relationships do constitute weak ties then other theorized network properties should also be manifest, namely homophily and triadic closures. Our findings support these theoretical conjectures. Firstly, results indicate that the strength of a tie is directly proportional to the level of similarity between users in the network (homophily property). Secondly, we found strong support for the triadic closure property as well; we developed a computational model to predict the formation of weak ties via triadic closures with an accuracy of 97.8%. Insights from our model can be used to improve a collaborative filtering approach for resource recommendation by predicting future similarity between users in the network.