This paper seeks to explore the connection between segregation in digital and physical spaces. We perform an exploratory analysis using a sample of geocoded Twitter data for the metropolitan area of Chicago to address two specific questions:
- Are the spatial mobility patterns of Twitter users and their tweets similar to the patterns of physical segregation along the lines of race, income, and education?
- Are Twitter users’ friend networks related to the demographic characteristics of the neighborhoods in which they reside?
Our results suggest that although the spatial movements of Twitter users tend to align with known residential segregation patterns in Chicago, the friend networks of these users do not seem to be related to these demographic patterns in any obvious way. We also use descriptive statistics to investigate the potential bias in our Twitter sample, and find that, on average, higher-income, highly educated, and whiter neighborhoods tend to be overrepresented.