Skip to main content
. 2015 Apr;105(4):732–740. doi: 10.2105/AJPH.2014.302374

TABLE 1—

Nested Models Showing the Associations Between Predictors and Depressive Symptoms: National Longitudinal Study of Adolescent Health, 1994–1995

Model 1
School-Only MLM Neighborhood-Only MLM CCMM Model 2: CCMM
Fixed-effect estimates
Intercept (SE) 11.2* (0.09) 11.1* (0.08) 10.8* (0.14) 3.6* (0.61)
Individual-level predictor
 Age 0.40* (0.32, 0.47)
 Female 1.97* (1.74, 2.21)
 Public assistance 1.70* (1.31, 2.09)
 College degree (parent) −1.46* (-1.72, −1.20)
 White (Ref) 1.00
 Black 1.00* (0.66, 1.35)
 Hispanic 1.52* (1.13, 1.91)
Random-effect estimates
Neighborhood (95% CI) 1.84* (1.32, 2.37) 0.45* (0.10, 0.79) 0.19* (-0.02, 0.39)
School (95% CI) 2.05* (1.5, 2.61) 1.69* (1.10, 2.28) 0.70* (0.37, 1.02)
Individual (95% CI) 54.65* (53.43, 55.86) 55.04* (53.81, 56.27) 54.68* (53.48, 55.89) 52.81* (51.64, 53.99)
Fit statistics
Deviance information criteriona 110 788 110 169
Intraclass correlation coefficientb estimates, %
Neighborhood 3.2 0.8 0.4
School 3.6 3.0 1.3

Note. CCMM = cross-classified multilevel model; CI = credible interval; MLM = multilevel model. The sample size was n = 16 172. For fixed-effect estimates, entries are parameter estimates (b) and credible intervals. All other entries are estimates (of variance) and credible intervals.

a

A measure of model fit reported only for cross-classified multilevel models. Higher values indicate a poorer fitting model.

b

Proportion of variance in the outcome attributable to neighborhoods (after adjustment for schools) or to schools (after adjustment for neighborhoods).

*P < .05.