Table 1.
Evaluation of agreement between residential density calculations and actual residential density determined by e911 data
| Residential Density Calculation | R-squared# | MSE@ | Slope# | (95% CI) | Intercept# | (95% CI) |
|---|---|---|---|---|---|---|
| Census Geography | ||||||
| Census tract | 0.63 | 0.32 | 0.49 | (0.45, 0.53) | 0.59 | (0.46, 0.71) |
| Census block group | 0.66 | 0.29 | 0.44 | (0.41, 0.48) | 0.36 | (0.26, 0.46) |
| Census block | 0.61 | 0.34 | 0.41 | (0.38, 0.45) | 0.04 | (-0.05, 0.13) |
| GIS-refined Measure | ||||||
| 1-km circle buffer | 0.89 | 0.09 | 0.59 | (0.57, 0.61) | 0.21 | (0.17, 0.26) |
| 1-km circle and road network buffer | 0.90 | 0.08 | 0.75 | (0.73, 0.78) | -0.11 | (-0.11, -0.07) |
#R-squared, slope, and intercept are from models in which both the calculated residential density and e911 residential density are transformed to the log scale.
@MSE is the variance estimate for the difference between calculated residential densities and e911 residential density; it reflects both model error and lack of fit from the identity line as illustrated in Figure 1.