When the clock chimes: The impact of on-the-hour effects on user anxiety content generation in social media platforms

Background: The growth of social media platforms has created a plethora of user-generated content, and social media has become an important channel of users to express their emotions. Although many studies have explored the influencing factors on user-generated content, there is an insufficient understanding the impact temporal cues on mental health content generation. Objective: This study aimed to explore how the on-the-hour time points affect users' anxiety content generation on social media platforms. Further, this study investigates the difference between weekdays and weekends, and the moderating effects of regional economic levels. Methods: We collected information on 2,543,902 user-generated anxiety-related posts from a leading social media platform in China. Then, we used fixed effect models to analyze the relationship between on-the-hour time points and user anxiety content generation. Results: The results show that on-the-hour time points affect user anxiety-related content generation, especially at the beginning of each hour (β = 894.564, p < 0.01). And the impact is greater on weekdays (β = 774.695, p < 0.01) than on weekends (β = 119.869, p < 0.01). Furthermore, regional economic moderates the impact, the better the economic condition, the greater the impact. Limitations: Limitations include incomplete coverage of patient types and a single anxiety dictionary. Conclusions: This study uncovers the relationship between temporal cues and user-generated anxiety content, providing new insights into the mental illness observation, and provides insights for mental health services providers and designers of online social platforms.