Abstract
This study analyzed 5.3 million tweets with #Wuhan and #Chinesevirus collected from X (formerly Twitter) API to enhance understanding of cross-cultural communications on social media. Using the Affect-as-information (AAI) model and the Latent Process model (LPM), cross-correlational time series analyses revealed significant links between real-world affective events and the expression of negative emotions on X. At the beginning of the coronavirus outbreak in Wuhan, a significant cross-correlation emerged between the number of newly hospitalized COVID-19 patients and negative emotions such as anger, fear, disgust, and sadness expressed on X. However, this correlation weakened as the pandemic progressed, though the intensity of fear remained stable. Additionally, the study identified an inverse relationship in expressed negative emotions between users with slight and strong political biases over time. Specifically, as the pandemic advanced, the difference in emotional expression between these two groups became less pronounced. These findings contribute empirical evidence of emotional arousal's influence on online discourse and address a critical gap in the literature by applying the AAI and LPM frameworks to computer-mediated cross-cultural communication. Overall, the results illustrate how emotionally charged real-world stimuli shape online sentiment, reinforcing the interconnectedness of digital and offline sociopolitical dynamics during a global public health crisis.