Table 5.

Multilevel mixed-effects logistic regression predicting foreign-born homicide victimization using a 2017 NVDRS sample (N = 8,597).

		b (se)	OR
Victim characteristics
Age	0–14	ref	ref
	15–24	1.59 (0.51)**	4.91
	25–34	2.02 (0.51)**	7.49
	35–54	2.26 (0.51)**	9.61
	55 and up	2.63 (0.51)**	13.82
Sex (male)		−0.06 (0.14)	0.94
Married		0.84 (0.12)**	2.30
Race/Ethnicity	White	ref	ref
	Black	−0.20 (0.17)	0.82
	Asian/Pacific Islander	4.04 (0.24)**	57.02
	Other/unspecified	0.62 (0.33)	1.86
	Hispanic	2.84 (0.15)**	17.16
Education	Below high school	ref	ref
	High school degree/GED	−0.42 (0.11)**	0.66
	Some college credit	−0.54 (0.18)**	0.58
	Associate degree	−0.36 (0.25)	0.70
	Bachelor degree/higher	0.34 (0.20)	1.40
Mental health diagnosis		−0.62 (0.30)*	0.54
Alcohol problem		−0.17 (0.29)	0.85
Substance abuse problem		−0.38 (0.17)*	0.68
Incident/circumstance characteristics
Victim–suspect relationship	Current or former partner	ref	ref
	Family member	−0.42 (0.25)	0.66
	Acquaintance	−0.15 (0.21)	0.86
	Stranger	0.22 (0.24)	1.24
	Other	−1.23 (1.11)	0.29
	Unknown	0.02 (0.20)	1.02
Method/weapon	Firearm	ref	ref
	Sharp instrument	0.69 (0.14)**	2.00
	Blunt object	0.05 (0.24)	1.05
	Hanging, strangulation	0.75 (0.27)**	2.11
	Other	0.30 (0.19)	1.35
Location of injury	House, apartment	ref	ref
	Street, sidewalk, alley	0.18 (0.14)	1.19
	Public use area	0.52 (0.17)**	1.68
	Other	0.29 (0.14)*	1.33
Victim used a weapon		−0.40 (0.25)	0.67
Gang-related		−0.55 (0.18)**	0.58
Homicide precipitated by another crime		0.15 (0.13)	1.16

Groups = 33 Wald chi = 918.14** State variance = 6.39% R-squared from single-level model = 0.32.

p < .01 = ** p < .05 = * ref = reference category.

Missing: Married (n = 115), Race (n = 3), Education (n = 173), Method/weapon (n = 209), Location of injury (n = 486).