Novel developments in network analysis for personality research

Network models have been recently proposed as a novel theoretical approach to personality. We present several strategies to define personality networks and show how network indices, such as centrality and clustering coefficients, can offer insights into the structure of personality. We focus especially on the clustering coefficients, for which we propose new adaptations that guarantee their interpretability in the context of personality networks. Moreover, we present strategies to identify clusters of nodes, not only relying on their mutual connections, but also considering the context provided by the entire network. Such clusters consist of nodes that are functionally equivalent (or topologically overlapped) and that can be merged together into a single node before performing network analysis. The performances of the new techniques and of the new indices are showed both using simulated networks and with real data.