Pakistan is one of the few remaining countries where wild poliovirus remains endemic, despite decades of eradication campaigns. Yet, vaccine hesitancy persists, not merely due to biomedical skepticism but through digital discourse. Drawing on 6,399 Urdu language tweets, this study uses natural language processing and lexicon based modeling to test four hypotheses on the emotional and symbolic drivers of hesitancy. Emotions are operationalized using the NRC Emotion Lexicon, treating trust and fear as measurable affective signals. The findings challenge Western behavioral models such as the Extended Parallel Process Model (EPPM): trust is negatively associated with hesitancy, whereas fear is positively associated with it even when trust is present. Notably, fear’s effect weakens in security framed tweets, which express moral resolve and collective strength rather than panic. Religious framing also predicts hesitancy, but it is often based on misquoted or misinterpreted religious references. In many cases, such discourse misaligns with the actual teachings of the religion, which historically endorse disease prevention and public health. Vaccine hesitancy in this context emerges not as an individual risk judgment, but as a culturally embedded form of communicative resistance, requiring discourse based, context sensitive approaches to global health communication.
Muhammad Awais (Mon,) studied this question.