Examining the Cultural Influence on Online Stances Towards COVID-19 Preventive Measures and Their Impact on Incidence and Mortality: A Global Stance Detection Analysis of Tweets

During the COVID-19 pandemic, nations implemented various preventive measures, triggering varying online responses. This study examines cultural influences on public online stances toward these measures and their impacts on COVID-19 cases/deaths. Stance detection analysis was used to analyze 16,428,557 Tweets regarding COVID-19 preventive measures from 95 countries, selected based on Hofstede’s cultural dimensions. To ensure the variety of population, countries were chosen based on Twitter data availability and a minimum sample size of 385 tweets, achieving a 95% confidence level with a 5% margin of error. The weighted regression analysis revealed that the relationship between culture and online stances depends on the cultural congruence of each measure. Specifically, power distance positively predicted stances for all measures, while indulgence had a negative effect overall. Effects of other cultural indices varied across measures. Individualism negatively affected face coverings stances. Uncertainty avoidance influenced lockdown and vaccination stances negatively but had a positive effect on social distancing stances. Long-term orientation negatively affected lockdown and social distancing stances but positively influenced quarantine stances. Cultural tightness only negatively affected face coverings and quarantine stances. Online stances toward face coverings mediated the relationship between cultural indices and COVID-19 cases/deaths. As such, public health officials should consider cultural profiles and use culturally congruent communication strategies when implementing preventive measures for future pandemics. Furthermore, leveraging digital tools is vital in navigating and shaping online stances to enhance the effectiveness of these measures.