Table 2.
Parameter estimates (SEs) for two-regime linear regression and spatial error models.
| Variablea | Linear regression model | Spatial error model |
|---|---|---|
| New York City regime | ||
| Intercept | 0.8147 (0.0960) | 0.8650 (0.1232) |
| Percent built < 1940 | 0.0110 (0.0017) | 0.0102 (0.0018) |
| Percent without high school diploma | 0.0059 (0.0024) | 0.0060 (0.0030) |
| Percent African American | 0.0069 (0.0012) | 0.0060 (0.0014) |
| Upstate/Long Island regime | ||
| Intercept | 0.2926 (0.0414) | 0.3675 (0.0521) |
| Percent built < 1940 | 0.0244 (0.0010) | 0.0233 (0.0011) |
| Percent without high school diploma | 0.0182 (0.0022) | 0.0153 (0.0023) |
| Percent African American | 0.0091 (0.0011) | 0.0108 (0.0013) |
| ρ | NA | 0.4649 (0.0383) |
| r2 | 0.63 | 0.52b |
NA, not applicable.
a
Dependent variable Y is ln(% elevated BLLs + 1) for both models.
b
Pseudo r2 (Buse adjustment) used for spatial error model.