The study explores the identification of highly influential nodes in the complicated grief (CG) network using network analysis. The CG network is conceptualized as a set of mutually reinforcing symptoms, and the study aims to determine which symptoms are most influential in the development and persistence of CG. The researchers developed two new indices, expected influence (EI), to account for both positive and negative edges in the network, which are not captured by traditional centrality measures. They evaluated these indices in simulated networks with both positive and negative edges and in an empirical CG network derived from a longitudinal study of bereavement. In simulated networks with exclusively positive edges, both centrality and EI indices were strongly associated with observed node influence. However, in networks with negative edges, EI indices were more strongly associated with observed influence than centrality indices. In the empirical CG network, both centrality and EI indices were correlated with the strength of the association between node change and network change. The study found that nodes with high expected influence, such as emotional pain and feeling of emptiness, were more strongly associated with changes in the overall CG network than nodes with low expected influence. The study highlights the importance of considering both the position of a node within the network and the nature of its relationships with other nodes. The proposed EI indices, which account for both positive and negative edges, may be more effective in identifying influential nodes in networks with negative edges. The findings suggest that high-EI nodes may be especially important to the etiology and treatment of CG. However, the study also notes limitations, including the potential overestimation of correlations due to the exclusion of external influences and self-loops in the simulations, and the possibility of omitting critical nodes in the CG network due to the use of DSM-5 diagnostic criteria. The study emphasizes the need for further research to better understand the dynamics of CG networks and to evaluate the effectiveness of interventions targeting specific nodes or edges in the network.The study explores the identification of highly influential nodes in the complicated grief (CG) network using network analysis. The CG network is conceptualized as a set of mutually reinforcing symptoms, and the study aims to determine which symptoms are most influential in the development and persistence of CG. The researchers developed two new indices, expected influence (EI), to account for both positive and negative edges in the network, which are not captured by traditional centrality measures. They evaluated these indices in simulated networks with both positive and negative edges and in an empirical CG network derived from a longitudinal study of bereavement. In simulated networks with exclusively positive edges, both centrality and EI indices were strongly associated with observed node influence. However, in networks with negative edges, EI indices were more strongly associated with observed influence than centrality indices. In the empirical CG network, both centrality and EI indices were correlated with the strength of the association between node change and network change. The study found that nodes with high expected influence, such as emotional pain and feeling of emptiness, were more strongly associated with changes in the overall CG network than nodes with low expected influence. The study highlights the importance of considering both the position of a node within the network and the nature of its relationships with other nodes. The proposed EI indices, which account for both positive and negative edges, may be more effective in identifying influential nodes in networks with negative edges. The findings suggest that high-EI nodes may be especially important to the etiology and treatment of CG. However, the study also notes limitations, including the potential overestimation of correlations due to the exclusion of external influences and self-loops in the simulations, and the possibility of omitting critical nodes in the CG network due to the use of DSM-5 diagnostic criteria. The study emphasizes the need for further research to better understand the dynamics of CG networks and to evaluate the effectiveness of interventions targeting specific nodes or edges in the network.