Table 3.
Multi-Level Logit Models Predicting High School Graduation
| Combined | Males | Females | ||||
|---|---|---|---|---|---|---|
| (1) | (2) | (3) | (4) | (5) | (6) | |
| Neighborhood Disadvantage Scale | −.142* (2.582) |
−.036 (.654) |
−.176* (.060) |
−.068 (1.115) |
−.128 (1.855) |
−.007 (.096) |
| Neighborhood Violence Scale | -- | −.152* (4.343) |
-- | −.149* (4.139) |
-- | −.169* (4.024) |
| Constant | 1.042* (8.141) |
.936* (7.038) |
1.115* (7.637) |
1.045* (6.967) |
1.320* (8.919) |
1.209* (7.954) |
| Variance Components: | ||||||
| School | .142 | .129 | .041 | .039 | .160 | .149 |
| Neighborhood | .064 | .035 | .035 | .004 | .115 | .088 |
| N individuals | 14,668 | 14,668 | 6,936 | 6,936 | 7,732 | 7,732 |
Notes:
All models include Individual Violence Scale, Neighborhood Social Cohesion and Intergenerational Closure Scales, and Individual, Family, and School Controls. Coefficients on these control variables available in Appendix B.
Absolute value of t-ratios in parentheses.
Multiple Imputation with 10 imputations used to impute missing values on control variables.
*
p < .05