Skip to main content
. 2014 Nov;134(5):e1362–e1368. doi: 10.1542/peds.2014-1395

TABLE 2.

Multilevel Poisson Regression Models Using AL at Age 19 as the Outcome

Model 1a Model 2a Model 3a Model 4b
B SE P B SE P B SE P B SE P
Constant 0.21 0.42 .63 0.26 0.43 .55 0.21 0.43 .63 0.37 0.35 .29
Between-neighborhood
 Neighborhood poverty (2000) −0.14 0.09 .12 −0.10 0.09 .30 −0.17 0.10 .09 −0.06 0.05 .21
 Neighborhood poverty (2010) 0.20 0.10 .05 0.17 0.10 .09 0.17 0.10 .10 0.08 0.05 .11
 NP (2000) × NP (2010) −0.12 0.06 .03 −0.19 0.06 .00 −0.11 0.04 .01
 ES × NP (2000) 0.20 0.09 .04 0.08 0.06 .17
 ES × NP (2010) −0.07 0.10 .49 0.03 0.05 .61
 NP (2000) × NP (2010) × ES 0.19 0.07 .01 0.10 0.05 .04
Between-person
 Family poverty at age 11 0.17 0.11 .13 0.13 0.18 .46 0.14 0.18 .45 0.05 0.15 .72
 Family poverty at age 19 0.03 0.12 .79 0.01 0.16 .93 0.02 0.16 .89 0.01 0.12 .93
 FP (age 11) × FP (age 19) 0.03 0.23 .90 0.04 0.23 .86 0.11 0.19 .57
 Emotional support at age 18 −0.02 0.06 .68 −0.02 0.05 .62
 Male 0.08 0.10 .44 0.10 0.11 .36 0.13 0.11 .24 0.11 0.09 .20
 Diet at age 19 −0.01 0.03 .89 −0.01 0.03 .98 0.01 0.03 .86 −0.01 0.02 .56
 Smoking at age 19 −0.08 0.13 .54 −0.08 0.13 .55 −0.03 0.13 .84 −0.07 0.11 .50
 Binge drinking at age 19 −0.08 0.14 .57 −0.03 0.14 .83 −0.05 0.14 .74 0.02 0.11 .86
 Perceived stress at age 19 −0.01 0.01 .69 −0.01 0.01 .70 −0.01 0.01 .61 −0.01 0.01 .62
 Unemployment at age 19 0.03 0.12 .78 0.02 0.12 .86 −0.03 0.12 .79 0.05 0.10 .64
 Financial stress at age 19 0.02 0.02 .40 0.02 0.02 .36 0.02 0.02 .23 0.01 0.02 .73
 Residential stability 0.03 0.09 .73
Random effect:
 τ (same census tract between 2000 and 2010) 3.72e-07 8.00e-10 3.51e-11
  τ (2000 census tract) 8.39e-06
  τ (2010 census tract) 4.1e-05
Deviance 884.96 879.85 863.44 1289.64

Neighborhood poverty and emotional support were standardized by z transformation. ES, emotional support; FP, family poverty; NP, neighborhood poverty.

a

Multilevel Poisson regression model with participants who evinced residential stability between 2000 and 2010, n(person) = 284 and n(neighborhood) = 41.

b

Cross-Classified Multilevel Poisson regression model with the inverse of a propensity score weighting with the entire sample, n(person) = 420, n(neighborhood 2000) = 48, and n(neighborhood 2010) = 88.