Skip to main content
. 2021 Oct 18;37(7):1057–1069. doi: 10.1007/s10896-021-00328-8

Table 3.

Logistic regression models of the B-SAFER victim vulnerability factors’ predictability of elevated summary risk ratings for severe or lethal re-victimization among female IPV victims with and without children, respectively

Children present (n = 457) β Wald p OR 95% CI
Modela
   Inconsistent attitudes or behavior 0.1 0.1 .806 1.3 [0.6, 2.5]
   Extreme fear of perpetrator 1.0 22.4 .000 3.9 [2.0, 7.7]
   Inadequate support or resources 0.5 4.4 .037 3.2 [1.4, 7.2]
   Unsafe living situation 0.4 2.7 .103 1.6 [0.7, 3.6]
   Health problems 0.2 0.8 .377 1.0 [0.5, 2.2]
Constant -1.6 51.7 .000
No children (n = 368) β Wald p OR 95% CI
Modelb
   Inconsistent attitudes or behavior 0.1 0.1 .810 1.1 [0.6, 1.9]
   Extreme fear of perpetrator 1.2 22.2 .000 3.2 [2.0, 5.3]
   Inadequate support or resources -0.1 0.1 .753 0.9 [0.5, 1.6]
   Unsafe living situation 1.2 18.4 .000 3.2 [1.9, 5.4]
   Health problems 0.5 3.1 .077 1.7 [0.9, 2.9]
Constant -2.0 60.6 .000

Notes. B-SAFER = Brief Spousal Assault Form for the Evaluation of Risk (Kroppet al., 2010). IPV = Intimate partner violence. OR = Odds ratios. CI = confidence interval. Summary risk ratings dichotomized as ‘low risk’ (i.e., low risk) and ‘elevated risk’ (i.e., moderate or high risk). Due to missing values, two sub-samples (n = 457 and n = 368) with complete coding for summary risk ratings of severe/lethal IPV in the B-SAFER were included for analyses

a Omnibus tests of model coefficients = χ2 (5) = 54.20, p = .000. Cox & Snell R square = .112. Nagelkerke R square = .152

b Omnibus tests of model coefficients = χ2 (5) = 75.94, p = .000. Cox & Snell R square = .186. Nagelkerke R square = .253