Use of Network Psychometrics Approach to Examine Social Media Disorder Symptoms

Abstract

The symptoms of behavioral addiction are generally regarded as a consequence of a latent construct. However, network psychometrics enable conceptualizing them as directly interacting with variables in a network of symptoms. This study, we aimed to investigate symptoms of social media disorder within this framework. This is the first study performed using this novel in the field of behavioral addiction, and conceptualizing social media disorder in this manner helps the professionals in gaining new insights on the construct. The data were collected by applying the Short Social Media Disorder Scale to 727 university students and were analyzed with qgraph and EstimateGroupNetwork packages in R program. Strength, closeness, and betweenness centrality indices were used to evaluate the most important symptoms in the network. The centrality of the network model was further investigated with Zhang's clustering coefficient and the small-world Index was calculated. Finally, the estimated network structures were compared based on gender and age variables. According to the results, withdrawal and preoccupation were detected as the most important symptoms, whereas deception was less important. In addition, it was found that the estimated network had a small-world property. These findings were discussed in terms of their theoretical and practical significance.