Table 1. Summary of the scaling exponents γ measured on the Twitter dataset.
| Geographical aggregation | Scaling exponent γ |
|---|---|
| Basins (internal connections only) | 1.11 ± 0.01 |
| Basins (all connections) | 1.11 ± 0.01 |
| Metro areas (internal connections only) | 1.20 ± 0.02 |
| Metro areas (all connections) | 1.08 ± 0.02 |
| Basins in US (internal connections only) | 1.15 ± 0.01 |
| Basins in US (all connections) | 1.16 ± 0.02 |
| Metro areas in US (internal connections only) | 1.16 ± 0.02 |
| Metro areas in US (all connections) | 1.09 ± 0.03 |
| Basins in Europe (internal connections only) | 1.21 ± 0.04 |
| Basins in Europe (all connections) | 1.06 ± 0.01 |
| Metro areas in Europe (internal connections only) | 1.18 ± 0.02 |
| Metro areas in Europe (all connections) | 1.08 ± 0.02 |
Error intervals correspond to the standard error of the slope in the regression fit. When referring to Europe, the following 31 countries are taken into consideration: Belgium, France, Bulgaria, Bosnia Herzegovina, Croatia, Germany, Hungary, Finland, Denmark, Netherlands, Portugal, Latvia, Lithuania, Luxembourg, Romania, Poland, Greece, Estonia, Italy, Albania, Czech Republic, Cyprus, Austria, Ireland, Spain, Macedonia, Slovakia, Malta, Slovenia, United Kingdom, Sweden.