In this paper, we analyze more than 16 million Twitter messages related to the 2022 war in Ukraine to examine the effects of multiple exposure to messages conveying intense anxiety or positivity on Twitter user behavior. We first analyzed a data-set covering a 3-month pre-war period to derive baseline anxiety and positivity levels. Subsequently, we compared the anxiety and positivity levels during the first 3 months after the war started in relation to the baseline. Our analysis indicates that the initial multi-exposure to intense anxiety is subsequently associated with a weaker expression of positivity as compared to users who initially have predominantly been exposed to positive messages. Moreover, anxiety-exposed users exhibit anxiety levels higher than their baseline in the post-exposure phase (i.e. during the second and third month of the war). In contrast, positivity-exposed users consistently show higher intensity of positivity and do not cross their baseline level of anxiety in the post-exposure phase. The low levels of positivity after an initial exposure to intense anxiety point to potentially disruptive mid-term effects of anxiety-conveying messages. Moreover, our results also point to the undoing effects that positive messages had in the early stages of the war.

Top