Table 1A.
Mediation models. Unstandardized beta coefficients (β) and 95% confidence intervals (CI) from mediational models calculated using Heyes Process bias corrected bootstrap method (5000 samples). *p < .05, **p < .01, ***p < .001. CTQ = Childhood Trauma Questionnaire, abuse subscales. A-COPE = Adolescent Coping Orientation for Problem Experiences, bespoke subscale assessing drug, smoking and alcohol use to cope. AUDIT = Alcohol Use Disorders Identification Test. DUDIT = Drug Use Disorders Identification Test. The a, b, and c paths indicate that there were significant interrelationships between CTQ abuse types (emotional, physical, sexual), A-COPE drug use coping motives, and alcohol/drug problems indexed by the AUDIT/DUDIT. The one exception is that CTQ sexual abuse was not related to drug use problems in the DUDIT (c path). All six indirect mediational paths were significant (the 95% confidence intervals do not encompass zero) demonstrating that coping motives mediate the relationship between childhood abuse and alcohol/drug problems. The direct effect (c' paths) between X→Y remained significant when the mediational paths were controlled for in four models, indicating partial mediation, whereas the direct effect was not significant for models with sexual abuse as the predictor indicating full mediation. The overall conclusion is that drug use coping motives mediate the relationship between childhood abuse and alcohol/drug problems in school attending South African adolescents.
| Component paths |
Mediation |
||||||
|---|---|---|---|---|---|---|---|
| X (predictor) | M (mediator) | Y (outcome) |
a path: X→M |
b path: M→Y controlling for X |
c path: total effect X→Y |
c' path: direct effect X→Y controlling for the indirect mediational path |
Indirect mediational path: X→M→Y |
| CTQ Emotional abuse score | A-COPE drug use coping motives | AUDIT total | β = .04, CI = .03- .05*** | β = 4.09, CI = 3.67-4.52*** | β = .26, CI = .17-.35*** | β = .09, CI = .02-.17* | β = .17, CI = .12-.22*** |
| CTQ Physical abuse score | β = .04, CI = .02- .05*** | β = 4.12, CI = 3.70-4.54*** | β = .24, CI = .15-.33*** | β = .10, CI = .02-.18* | β = .14, CI = .09-.20 *** | ||
| CTQ Sexual abuse score | β = .05, CI = .03- .06*** | β = 4.17, CI = 3.74-4.59*** | β = .24, CI = .13-.34*** | β = .04, CI=-.05-.13 | β = .19, CI = .13-.26 *** | ||
| CTQ Emotional abuse score | DUDIT total | β = .04, CI = .03- .05*** | β = 2.56, CI = 2.25-2.87*** | β = .19, CI = .13-.25*** | β = .09, CI = .03-.14* | β = .10, CI = .07-.14 *** | |
| CTQ Physical abuse score | β = .04, CI = .02- .05*** | β = 2.58, CI = 2.27-2.89*** | β = .18, CI = .11-.24*** | β = .09, CI = .03-.14** | β = .09, CI = .06-.13 *** | ||
| CTQ Sexual abuse score | β = .05, CI = .03- .06*** | β = 2.72, CI = 2.41-3.03*** | β = .06, CI = .-01-.14 | β=-.06, CI=-.13-.01 | β = .13, CI = .08-.18*** | ||