Table 3.

Effects of the intervention (outcomes for the intervention versus control group)

	Unadjusted means				Regression-adjusteda,b,c
	Control	Intervention	Difference	p	Effect	p
	(N = 29)	(N = 31)
Outcomes on Likert/rating scales
Readiness: seek information (0–10)	3.52	5.35	1.84	.050	1.60	.100
Readiness: talk to family (0–10)	2.38	5.35	2.98	.002	2.74	.007
Readiness: talk to friend (0–10)	2.55	5.35	2.80	.002	2.48	.010
Readiness: see MH professional (0–10)	2.45	5.71	3.26	.000	3.16	.001
Readiness: self-help or support group (0–10)	1.10	1.87	0.77	.240	0.50	.460
Readiness: academic support services (0–10)	2.07	1.68	−0.39	.550	−0.44	.520
Readiness: AVERAGE (0–10)	2.34	4.22	1.88	.002	1.67	.008
Perceived public stigma score (0–4)	3.21	2.71	−0.50	.080	−0.59	.040
Personal stigma score (0–4)	2.00	0.90	1.10	.003	−1.07	.004
PHQ-9 score (0–27)	11.0	11.39	0.39	.755	0.29	.811
AUDIT score (0–40)	6.00	6.84	0.84	.600	0.51	.527
Binary outcomes
Perceived a need for help (2 mos.)	62%	72%	0.10	.660	OR = 1.51	.480
Met w/ MH professional (2 mos.)	0%	28%	0.28	.002	**
Therapy/counseling (current)	0%	13%	0.13	.050	**
Psychiatric medication (current)	0%	9%	0.09	.090	**

Note. OR = Odds Ratio

^aLinear regressions were estimated for Likert/rating scale outcomes, logistic regressions for binary outcomes

^bCovariates include: age, gender, and baseline suicide risk factors (PHQ-9 score, AUDIT score, suicidal ideation or previous attempt).

^cIn a Bonferroni adjustment for multiple hypothesis testing, we have nine outcomes (with average readiness a composite outcome); therefore, the unadjusted p-values shown above would only be significant at the standard 0.05 level if they are less than 0.0056. With this adjustment, effects remain significant for average readiness, met with a mental health professional, and personal stigma.

^**Cannot be estimated because no variation in outcome in control group.