Testing the Empirical Foundation for Secondary Prevention of Family Maltreatment in the U.S. Air Force

Background and Purpose: The United States Air Force (USAF) has implemented secondary prevention activities aimed at preventing family maltreatment among active duty members and their families. In this context, USAF Family Advocacy Program (FAP) leaders are asking whether the current array and combination of interventions are optimal for the secondary prevention of family maltreatment. Building on a logic model previously guiding service selection, a systematic review was conducted to identify key modifiable mediators empirically associated with the perpetration of family maltreatment among active duty members. Findings from the systematic review led to the formulation of a revised logic model, in which targets of future FAP secondary prevention efforts were proposed and explicated. This study empirically examines the revised logic model and its interrelated components via secondary data analysis of a large, probability sample of active duty Air Force members.

Methods: Source of Data: Data for the study came from the 2011 Air Force Community Assessment, a community assessment with over 300 survey items. Sample: The final analytical sample included 30,541 active duty Air Force members (83% male) who had at least one child and were in a committed couple relationship. Measures: Measures were selected based on their congruence with the revised logic model. The focal latent endogenous construct was personal coping (an asset-based proxy for maltreatment perpetration). The remaining constructs were individual fitness (second-order factor representing first-order factors for financial fitness, physical fitness, and psychological fitness); safe, stable, and nurturing family (SSNF; a second-order factor representing first-order factors for family coping, couple relationship quality, partner support, and parent-child relationship quality); informal support (second-order factor representing first-order factors for sense of community, neighbor support, and personal network support), and unit leader support. A grouping variable for neighborhood safety was also used to identify participants residing in unsafe versus safe neighborhoods. Analysis: Structural equation modeling with multiple-group comparison analysis was used to specify latent constructs, handle measurement error, analyze hypothesized structural parameters, and assess measurement and structural invariance between participants residing in safe and unsafe neighborhoods.

Results: The final model yielded acceptable fit (CFI = .97; TLI = .97; RMSEA = .02) and measurement invariance was established between neighborhood groups. After controlling for socio-demographic characteristics, results indicated that unit leader support was positively associated with informal support, individual fitness, and SSNF; Informal support and SSNF were positively associated with individual fitness; and informal support, SSNF, and individual fitness were positively associated with personal coping. The model explained more variance in endogenous constructs among those in unsafe neighborhoods compared to those in safe neighborhoods.

Conclusions and Implications: Overall, the results support and validate the revised logic, which points to key modifiable mediators across several socio-ecological levels that should be targeted by FAP activities, services, and programs to prevent family maltreatment. Specifically, FAP services should include evidence-based programs that attempt to bolster unit leader support, informal support, features of SSNF, individual fitness, and personal coping. Intervention efforts to prevent maltreatment may be particularly impactful among those who reside in neighborhoods perceived as unsafe.