فاطمه پورغلامی

Social media platforms are vital repositories of user-generated content, reflecting a range of emotions, interests, and discussions. Among these interactions, hate speech has emerged as a significant issue, influencing user behavior. While prior studies have attempted to analyze user characteristics to understand hate attitudes, they often rely on simple statistical comparisons and lack robust methods for causal effect estimation. This study investigates the causal effects of hate speech on user behavior on Twitter (now known as X) during the COVID-19 pandemic, characterized by heightened online discourse and harmful rhetoric. We focus on users who broadcast hate speech to determine how such expressions affect emotional responses. Using a Bayesian structural time-series modeling approach, we isolate the effects of hate speech from confounding factors, providing a solid framework for causal inference. Our findings indicate a significant shift in user emotions following instances of hate speech, demonstrating a measurable impact on user dynamics. We also analyze hashtag usage during this period, emphasizing their role in shaping online discourse. This study enhances understanding of the relationship between hate speech and user behavior, offering insights crucial for researchers, policymakers, and social media platforms in developing strategies to mitigate the adverse effects of online hate speech.