Table 3.
Results of Two-Level Logistic Regression Analysis
| Estimates | S.E. | Critical Ratio | Odds Ratio | |
|---|---|---|---|---|
| Within Level: | ||||
| STD on cocaine test results | 0.642 | 0.363 | 1.770 | 1.900 |
| Marijuana test results | 0.328 | 0.263 | 1.250 | 1.388 |
| Gender (male) | -1.144 | 0.201 | -5.690*** | 0.319 |
| Age | 0.330 | 0.086 | 3.828*** | 1.390 |
| Race (African American) | 1.413 | 0.255 | 5.548*** | 4.109 |
| Charge level (DJJ case) | 0.810 | 0.209 | 3.868*** | 2.248 |
| Between Level: | ||||
| Concentrated disadvantage by | ||||
| African American | 1.000 | -- | -- | |
| Female-headed with kids | 1.156 | 0.111 | 10.388*** | |
| Unemployed | 0.988 | 0.134 | 7.392*** | |
| Below poverty | 1.487 | 0.134 | 11.100*** | |
| STD on concentrated disadvantage | 1.561 | 0.795 | 1.965* | |
| Youth | -2.563 | 2.107 | -1.156 | |
| Residential stability | 1.192 | 1.032 | 1.156 | |
| Hispanic | 0.167 | 1.109 | 0.151 | |
| Ethnic heterogeneity | -0.760 | 0.887 | -0.856 | |
| Threshold for STD | 8.372 | 1.674 | 5.084*** | |
| Variance for concentrated disadvantage | 0.032 | 0.006 | 5.425*** | |
| Residual variance for | ||||
| STD | 0.190 | 0.189 | 1.006 | |
| African American | 0.019 | 0.003 | 5.793*** | |
| Female-headed with kids | 0.026 | 0.005 | 5.173*** | |
| Unemployed | 0.047 | 0.018 | 2.657** | |
| Below poverty | 0.032 | 0.005 | 6.344*** |
Note. Due to the estimation of random effects, calculation of odd ratios are not appropriate at the between level.
*
Two-tailed p-values: p < .05
**
p < .01
***
p < .001