Table 4.
Bayesian conditional autoregressive Poisson models for counts of violent crimes 2013–2014; 1940 Census tracts, Philadelphia (n = 404)
| Model 2a | Model 2b | Model 2c | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| IRR | (95% | CI) | IRR | (95% | CI) | IRR | (95% | CI) | |||
| HOLC Zone | |||||||||||
| Green (reference) | |||||||||||
| Blue (100% increase) | 2.711 | 1.282 | 6.534 | 1.863 | 0.718 | 6.639 | |||||
| Yellow (100% increase) | 4.225 | 1.683 | 10.392 | 2.415 | 0.800 | 8.882 | |||||
| Red (100% increase) | 5.518 | 2.466 | 14.325 | 2.571 | 0.913 | 11.001 | |||||
| No Zone (100% increase) | 10.665 | 4.568 | 26.128 | 4.531 | 1.565 | 17.288 | |||||
| 1940 Census Variables | |||||||||||
| White and Other Race (reference) | |||||||||||
| Black (10% increase) | 0.981 | 0.902 | 1.069 | 0.995 | 0.916 | 1.079 | |||||
| House value ($1,000 increase) | 0.906 | 0.852 | 0.955 | 0.946 | 0.888 | 1.001 | |||||
| Concentrated Disadvantage (z-score) | 1.451 | 1.199 | 1.774 | 1.354 | 1.124 | 1.656 | |||||
| Proportion variance explained by spatial random effect | 0.925 | 0.766 | 0.999 | 0.960 | 0.859 | 0.998 | 0.959 | 0.842 | 0.997 | ||
Nb. Bolding denotes credible intervals do not include the null value of 1.000, providing evidence of an association