Skip to main content
. Author manuscript; available in PMC: 2016 May 1.
Published in final edited form as: J Abnorm Psychol. 2015 Jan 19;124(2):329–342. doi: 10.1037/abn0000029

Table 3.

Concurrent and prospective relationships between CD/CU groups and outcomes

Omnibus test 1. No CD diagnosis mean n=215 2. CD – CU mean n=21 3. CD + CU mean n=14 Significant group differences
Concurrent relationships (age 17)
Aggressive symptoms F=146.4
p<.001
.07 (.28) 1.8 (1.5) 1.8 (1.3) 1<2, 3
Rule Breaking Symptoms F=99.1
p<.001
.40 (.71) 2.9 (1.3) 3.0 (1.8) 1<2, 3
Arrest record (adolescent) χ2=46.5
p<.001
24% (.43) 89% (.31) 92% (.29) 1<2,3
Prospective relationships (age 20)
Self-reported delinquency F=6.41
p<.01
10.6 (8.8) 19.6 (12.8) 11.1 (5.7) 1, 3<2
APD symptoms F=16.3
p<.001
.68 (1.4) 2.0 (1.5) 2.0 (1.9) 1<2,3
% APD diagnosis χ2=18.8
p<.001
13% (.34) 43% (.51) 43% (.51) 1<2,3
% Depression diagnosis χ2=11.1
p<.01
7% (.26) 29% (.46) 14% (.36) 1<2
% Substance Abuse diagnosis χ2=8.52
p<.05
25% (.44) 52% (.51) 43% (.51) 1<2,3
% Substance Dependence diagnosis χ2=17.8
p<.001
15% (.36) 52% (.51) 29% (.47) 1<2,3
CU traits (age 20) F=6.92
p=.001
2.4 (1.7) 2.9 (1.9) 4.2 (1.7) 1<3
Arrest record (adult) χ2=18.3
p<.001
26% (.44) 71% (.46) 29% (.47) 1,3<2

N=250 for all analyses and the data from the first imputed dataset are presented though note that results were equivalent across 5 imputed datasets. Statistical comparisons of the means were conducted on the log-transformed data (for aggressive, rule breaking and self-reported delinquency), as they better approximate normality. However, the corresponding raw symptom counts are presented in here and discussed in the text to promote ease of understanding. Post-hoc group differences were tested using a Tukey test (for ANOVAs with continuous outcomes) and individual χ2 tests (for binary outcomes) and are marked if p<.05. Results with binary outcomes are presented with χ2 tests but were similar in terms of significance when using Fisher exact tests or ANOVAs.