Understanding internal dynamics of state anxiety during COVID-19 pandemic: Seven-wave longitudinal findings via panel network analysis

Abstract

Research on state anxiety has long been dominated by the traditional psychometric approach that assumes anxiety symptoms have a common cause. Yet state anxiety can be conceptualized as a network system. In this study, we utilized data from the COVID-Dynamic dataset from waves 7 to 13, collected at three-week intervals from June 6, 2020, to October 13, 2020, and included 1,042 valid participants to characterize the internal dynamics of state anxiety. Using the Gaussian graphical model along with strength centrality, we estimated three network models of state anxiety. The between-subjects and contemporaneous network showed numerous positive relations between items and some unexpected negative relations. Three communities were identified in the between-subjects network, and two communities were identified in the contemporaneous network. The temporal network showed the coexistence of positive and negative predictions between items after three weeks. Several items exhibited significant positive autocorrelations after three weeks. These findings have implications for anxiety theory and clinical interventions at between-subjects and within-subjects levels.