Table 3.

Standardized Regression Coefficients: Indirect Effect Models

Y	Model	Model R2	β	p	Bootstrapped	95% CI
AUDIT	PCL-5 → DERS (a)	.18**	.37	<.001	.27	.46
	DERS → AUDIT (b)	.19**	.11	<.001	.06	.15
	PCL-5 → AUDIT (c)	.11*	.06	.001	.02	.09
	PCL-5 → AUDIT (c’)		.02	.30	-.02	.06
	PCL-5 → DERS → AUDIT (a*b)		.04		.02	.06
DMQ-R-SF: Coping motives subscale	PCL-5 → DERS (a)	.27**	.32	<.001	.22	.41
	DERS →DMQ-R-SF Coping motives subscale (b)	.39*	.02	.004	.01	.04
	PCL-5 → DMQ-R-SF Coping motives subscale (c)	.37*	.02	.001	.01	.03
	PCL-5 → DMQ-R-SF Coping motives subscale (c’)		.01	.04	.001	.03
	PCL-5 → DERS → DMQ-R-SF Coping motives subscale (a*b)		.01		.002	.01
DMQ-R-SF: Enhancement motives subscale	PCL-5 → DERS (a)	.29**	.29	<.001	.19	.39
	DERS → DMQ-R-SF Enhancement motives subscale (b)	.46	.01	.23	−.005	.02
	PCL-5 → DMQ-R-SF Enhancement motives subscale (c)	.46	.002	.65	−.008	.01
	PCL-5 → DMQ-R-SF Enhancement motives subscale (c’)		.0001	.98	−.01	.01
	PCL-5 → DERS → DMQ-R-SF Enhancement motives subscale (a*b)		.002		−.002	.01
DMQ-R-SF: Conformity motives subscale	PCL-5 → DERS (a)	.28**	.30	<.001	.20	.40
	DERS → DMQ-R-SF Conformity motives subscale (b)	.26*	.02	.008	.01	.04
	PCL-5 → DMQ-R-SF Conformity motives subscale (c)	.24	.01	.084	−.001	.02
	PCL-5 →DMQ-R-SF Conformity motives subscale (c’)		.004	.48	−.01	.02
	PCL-5 → DERS → DMQ-R-SF Conformity motives subscale (a*b)		.01		.002	.01
DMQ-R-SF: Social motives subscale	PCL-5 → DERS (a)	.30**	.29	<.001	.19	.38
	DERS → DMQ-R-SF Social motives subscale (b)	.38	−.0001	.99	−.02	.02
	PCL-5 → DMQ-R-SF Social motives subscale (c)	.38	−.01	.30	−.02	.01
	PCL-5 → DMQ-R-SF Social motives subscale (c’)		−.01	.33	−.02	.01
	PCL-5 → DERS → DMQ-R-SF Social motives subscale (a*b)		.00		−.005	.01

Note.

^**p < .001

^*p <.05.

The standard error and 95% CI for a*b are obtained using the bootstrapping technique with 5000 resamples. PCL-5 (PTSD symptom severity) is the predictor in all models. DERS (emotion regulation difficulties) is the indirect predictor in all models. CI (lower) is the lower bound of a 95% CI; CI (upper) is the upper bound of a 95% CI. Path a indicates effect of X on M; b, effect of M on Y controlling for X; c, effect of X on Y; c’, direct effect of X on Y, controlling for M. All total, direct, and indirect paths are noted after controlling for the variance accounted for by theoretically relevant covariates. Model 1 includes trauma load; the remaining models include trauma load and alternate alcohol use motives. PCL-5 = PTSD Checklist for DSM-5; DERS = Brief Difficulties in Emotion Regulation Scale; AUDIT = Alcohol Use Disorders Identification Test; DMQ-R-SF = Drinking Motives Questionnaire Revised Short Form.