The rural–urban stress divide: Obtaining geographical insights through Twitter.

To understand rural–urban differences in stressors, this study compared the cognitive and emotional language in geolocated Twitter posts in the United States against survey-reported county-level trends from the Gallup-Sharecare Well-Being Index. Mentions of stress on Twitter can predict population-level trends in stress in both rural (R2=31.6%) and urban (R2=26.7%) communities. While mentions of poor health are limited to only rural areas with low socioeconomic status, higher emotional expression is associated with higher stress across all rural communities. Controlling for socioeconomic status, urban communities reporting higher stress are also more likely to discuss relationships on Twitter. The findings contribute to an understanding of how language use on social media acts as a barometer of the social and cultural differences between regions. The data and stress topics developed in this paper are publicly available and can be accessed at https://osf.io/af8ce/. • Methods to derive data-driven dictionaries from social media, validated against survey data on community-level outcomes. • Social media mentions of stress discuss appraisal, problems, and coping processes aligning with psychological theory. • Controlling for socioeconomic status, rural areas with higher stress are more emotionally expressive on Twitter. • Controlling for socioeconomic status, urban areas with higher stress are more likely to discuss relationships on Twitter. [ABSTRACT FROM AUTHOR]