Post-9/11 Service Era Veterans: Intimate Partner Violence and Substance Use

Abstract

Using structural equation modeling, this study considers variations of intimate partner violence (IPV) among military families from the point of the perpetrator to test previously established empirical models on military subgroups in order to observe the impact of demographic factors on the type of IPV most prevalent among Post-9/11 military families from the National Longitudinal Study of Adolescent to Adult Health (1994–2008): Waves I and IV in-home interviews (N = 499). Study findings indicate that the perpetration of physical and sexual IPV varies across race/ethnicity perpetrator profiles. Models for substance use and IPV patterns were not similar across military cohorts and or racial/ethnic groups.

Keywords:

• IPV
• violence
• military
• veterans

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Statistically, structural equation modeling (SEM) allows one to operationalize a conceptual theoretical framework and its validation indices can verify several dimensions simultaneously measuring various aspects of intimate partner violence (IPV).

2 The calculated scores for these dimensions are measures that can serve as a basis for studying violence in the military family.

3 For the purpose of this study, men are considered not because they are assumed to be perpetrators, but because men are twice as likely to be stressed for an upcoming deployment than women, men are more likely to be stressed by their absence to their family than women, men are more likely to be stressed by financial reasons than women, men are more likely to cope with stress by drinking, and having violent obsessions and thoughts than women (Bray, Fairbank, & Marsden, 1999). According to the U.S. Department of Defense, most veterans and military personnel are men, 85% respectively (U.S. Department of Defense, 2015).

4 Some evidence exists that simple SEM models could be meaningfully tested even if the sample size is quite small (Hoyle, 1999; Hoyle & Kenny, 1999; Marsh & Hau, 1999), but usually, N = 100–150 is considered the minimum sample size for conducting SEM (Anderson & Gerbing, 1988; Tabachnick & Fidell, 2001; Tinsley & Tinsley, 1987), making both veteran samples adequate.

5 The strategy was to focus on solution propriety in lieu of trying to attend only to statistical power by taking the mean minimum sample size based on analyses with several different seed numbers

6 Alcohol and substance use variables were not factored into any latent constructs in order to explicitly identify differences between the various categories.

7 Goodness of fit was evaluated using three indices (Kline, 2011). The root means square error of approximation (RMSEA) incorporates a penalty function for poor model parsimony (Hu & Bentler, 1999). Values under 0.06 suggest close approximate (adequate) fit, whereas values above 0.10 signify a poorly fitting model and that the model should be rejected (Kline, 2011). The comparative fit index (CFI) and goodness-of-fit index (GFI) represent incremental fit indices contrasting the hypothesized theoretical model to a more restricted nested baseline model (Kline, 2011). Both range from zero to one and values >0.9 are indicative of adequate fit (Tucker & Lewis, 1973).

8 A modification index reflects how much the overall model chi-square decreases or improves if a constrained parameter is freely estimated. Candidate paths involving conspicuous values (modification indices > 10) were then examined for the actual amount of model fit improvement and for the magnitude of the following freed estimated coefficients. The decision to explore and keep new paths also followed their theoretical meaningfulness.

9 This process adheres to the hierarchical principle advocated in the theoretical model. The trimming process thus moved from left to right, starting with paths beginning from < Race/Ethnicity > to < Physical /Sexual>.