Skip to main content
. Author manuscript; available in PMC: 2020 Oct 1.
Published in final edited form as: Law Hum Behav. 2019 Aug 22;43(5):477–490. doi: 10.1037/lhb0000347

Table 4.

Fixed effect model estimates for the effect of change in total convictions on change in antisocial characteristics

Antisocial Attitudes Antisocial Behavior Antisocial Peers
β 95% CI p β 95% CI p β 95% CI p
Intercept 2.04 [1.34, 2.74] < .001 −0.36 [−1.28, 0.56] .432 −0.08 [−1.06, 0.91] .872
1-year lagged total convictions 0.03 [0.01, 0.05] .001 −0.01 [−0.03, 0.02] .509 0.03 [0.00, 0.06] .042
Age −0.12 [−0.12, −0.11] < .001 0.00 [−0.01, 0.01] .777 0.01 [−0.01, 0.02] .242
Race −0.09 [−1.06, 0.87] .842 0.25 [−0.97, 1.46] .677 0.24 [−1.01, 1.49] .691
2-year lagged
Alcohol use 0.10 [0.02, 0.18] .015 0.05 [−0.04, 0.15] .264 0.06 [−0.05, 0.18] .251
Marijuana use −0.01 [−0.09, 0.08] .829 0.01 [−0.12, 0.14] .867 −0.02 [−0.13, 0.08] .668
Academic achievement 0.06 [−0.02, 0.13] .095 0.03 [−0.05, 0.11] .472 −0.04 [−0.14, 0.07] .446
Antisocial attitudes 0.00 [0.00, 0.00] .184 0.00 [0.00, 0.01] .002 0.01 [0.00, 0.01] .005
Antisocial behaviors 0.00 [0.00, 0.00] .046 0.00 [0.00, 0.00] < .001 0.00 [0.00, 0.00] .529
Antisocial peers 0.02 [0.01, 0.02] < .001 0.01 [0.00, 0.02] .079 0.03 [0.02, 0.04] < .001
Internalizing 0.00 [−0.01, 0.00] .287 0.00 [0.00, 0.00] .947 0.00 [0.00, 0.00] .940
Neighborhood crime 0.00 [−0.01, 0.01] .987 0.00 [−0.01, 0.00] .441 0.00 [−0.01, 0.01] .991
Parent convictions −0.04 [−0.09, 0.01] .086 −0.01 [−0.07, 0.05] .703 −0.04 [−0.12, 0.05] .388
Parental stress 0.00 [−0.01, 0.01] .637 0.00 [−0.01, 0.01] .643 0.00 [−0.01, 0.01] .658
Socioeconomic status 0.01 [0.00, 0.01] < .001 0.00 [0.00, 0.00] .696 0.00 [−0.01, 0.01] .933
Parental supervision 0.00 [−0.02, 0.02] .960 0.01 [−0.03, 0.05] .674 −0.02 [−0.07, 0.02] .272
Psychiatric institutionalization 0.01 [−0.18, 0.2] .909 0.14 [−0.21, 0.49] .394 0.11 [−0.23, 0.45] .498
Delinquency institutionalization −0.03 [−0.07, 0.01] .122 −0.05 [−0.10, 0.00] .035 −0.06 [−0.20, 0.07] .320

Note. N = 503. CI: Confidence interval. The number of observations is 8,551. Coefficients for the 503 respondent dummy variables are not shown. Main effects are bolded, as the coefficients for confounding variables are not interpretable (Westreich & Greenland, 2013). Unadjusted models available upon request.