Table 4.

Moderation analysis results by population and measurement types.

	Suicide behaviors				NSSI
Correlate	k	Effect and 95% confidence			k	Effect and 95% confidence
		interval				interval
		Effect	Lower	Upper		Effect	Lower	Upper
Population type	Between-group heterogeneity				Between-group heterogeneity
	(Q = 2.63; df = 3, p = 0.000)				(Q = 2.39; df = 2, p = 0.303)
Clinical patients	1	−0.64	−0.79	−0.42	0	-	-	-
Community sample	7	−0.35	−0.42	−0.27	5	−0.31	−0.40	−0.21
Past suicide attempters	2	−0.17	−0.26	−0.07	1	−0.29	−0.49	−0.07
Marginalized sample	3	−0.36	−0.40	−0.32	1	−0.18	−0.31	−0.04
SCS measurement	Between-group heterogeneity				Between-group heterogeneity
	(Q = 0.80; df = 1, p = 0.372)				(Q = 0.50; df = 1, p = 0.479)
SCS long version	6	−0.30	−0.42	−0.19	4	−0.32	−0.44	−0.19
SCS short version	7	−0.36	−0.42	−0.31	3	−0.25	−0.40	−0.08
Suicide/NSSI measurement	Between-group heterogeneity				Between-group heterogeneity
	(Q = 1.35; df = 1, p = 0.245)				(Q = 0.00; df = 1, p = 0.980)
Screening items	3	−0.28	−0.38	−0.18	2	−0.29	−0.45	−0.11
Validated measure	10	−0.36	−0.42	−0.28	5	−0.29	−0.40	−0.17