Next Steps in Untangling the Web of Violence: A Research Agenda

Toward Evidence-Based Policy

The first important trend we would like to highlight is the shift to evidence-based policy (Pawson, 2006). The idea behind evidence-based policy is that we can bring the scientific method to laws and governmental regulations. The papers by Armstrong and Bosk and Victor and colleagues are good examples of this kind of work. Victor and colleagues (2021) analyzed child protective services manuals in 41 states and provided guidelines for substantiating children’s exposure to domestic violence and enhancing children’s and survivors’ safety and well-being. Armstrong and Bosk (2020) identified contradicting policy definitions and investigative procedures that allow workers to adopt different approaches in their decision-making, often leading to punitive approaches despite policy intentions. This study exemplifies the work necessary to successfully implement strong policies. Unfortunately, policy analysis remains one of the most understudied areas of violence research. Now is the time to start providing more evidence that can more explicitly guide policy makers and child welfare agency leaders.

Policy analyses are particularly needed on the interconnections among different forms of violence, including child abuse and domestic violence. Policies addressing the cumulative burden of trauma (trauma dosage) and the interconnections among forms of violence have been hampered by the siloed nature of many institutions, state laws, and regulations, which often focus on only one type (or closely related set of types) of violence. Social institutions need guidance to move toward integrated and comprehensive approaches to tackling family violence and trauma dosage.

In line with the focus of this special issue, we want to encourage people to start looking at the impact of policies on multiple outcomes. Any policy that impacts child maltreatment is likely to impact domestic violence and vice versa. Further, we want to encourage researchers to think beyond a focus on policies that have “child maltreatment” or “domestic violence” in the title. There is emerging data showing the benefits of an expanded social safety net and pro-family policies on violence. For instance, Berger and colleagues (2017) reported that an exogenous increase in income (via the earned income tax credit) was associated with reductions in neglect-related behaviors and CPS involvement, particularly in low-income single mothers. Conversely, Ginther and Johnson-Motoyama (2017) found that states who adopted restrictions in the Temporary Assistance to Needy Families (TANF) cash-assistance program saw an increased number of child maltreatment victims and foster care placements. These kinds of findings are potential gamechangers and we need more evidence on these and similar initiatives.

We also need studies that assess the impact of policies such as the Family First Prevention Services Act (FFPSA), which promotes the use of evidence-based prevention services listed on the Title IV-E Prevention Services Clearinghouse (https://preventionservices.abtsites.com/program). We also need evidence as to whether such lists adequately address the needs of system-involved children and caregivers. Although a large proportion of child welfare cases involve domestic violence, the FFPSA does not include domestic violence among its focus areas, which include parenting, substance abuse, and mental health. The FFPSA is a good example of an issue we highlight more below, that we need research on whether we can reduce the burden of trauma by supporting family strengths and reducing related problems. It is important policy makers collaborate with researchers and think more holistically about different forms of family violence when drafting policy, particularly policies that involve funding allocation for required services. Finally, we note that there can be unintended consequences to many policies, which are often formed in reaction to one problem (or even one tragedy) and turn out to affect others. We need more work on this, which could help explore how sometimes agencies that are all designed to serve the public can sometimes seem to be working at cross-purposes.

Many different types of studies can support this work. At the low-resource end, states, counties, and other jurisdictions often (largely unintentionally) create natural experiments. When one jurisdiction changes a law or policy, we can study the before and after. We can even compare them to similar nearby jurisdictions, further isolating the policy change as the cause of any impacts, in ways that mirror many elements of true experiments. Many states and governments engage in ongoing surveillance of family violence, crime, and related problems that essentially provides pretest and posttest data for such studies. Other organizations also collect relevant data. For example, the Robert Wood Johnson Foundation publishes annual data at the county level for a wide range of US health indicators. Rebbe and colleagues’ (2021) creative paper shows this kind of research in action, using population-based records to assess CPS responses to intimate partner violence reports. We encourage researchers to seek opportunities for partnerships between state agencies and research institutions and data sharing and coordination across agencies/institutions.

The challenges in such research are that the nature of the intervention and the types and quality of the data collected are largely out of researchers’ control, but the benefits are that they can be enormously influential in policy spheres that affect day-to-day lives. Because this is a neglected area of research, it is easier to make a notable contribution. Finally, although policy-focused research brings its own challenges, it is feasible to obtain such data in a relatively short period of time and thus would be suitable for graduate students, pre-tenure faculty, or others who are on a tight clock.

Examining the Impact of Trauma Dosage

Research on the interconnections of domestic violence and child abuse is part of a larger revolution on the impact of trauma dosage. Using frameworks such as adverse childhood experiences (ACEs) or poly-victimization (Felitti et al., 1998; Finkelhor et al., 2011), we now realize that the cumulative burden of trauma is the driver behind many psychological and physical outcomes (Hamby et al., 2021). One’s dosage of trauma in childhood appears to be especially important, as that developmental period impacts many maturing systems, for example, setting thresholds of cortisol reactivity. Childhood exposure may also be more likely to initiate vicious cycles leading to further adversity and trauma (Dierkhising et al., 2019). Cumulative trauma dosage is important because of the similarities in the body’s responses across different types of stressful events. Bessel van der Kolk’s memorable phrase, “the body keeps the score” (Van der Kolk, 1994) calls for rethinking much of our trauma dosage research and services.

In terms of research, at the low-resource end, one relatively accessible way of expanding research on trauma dosage and interconnections among violence and trauma is to use archived datasets. There are many important research questions that can be informed by existing data, including not only large, well-known datasets such as National Survey of Child and Adolescent Wellbeing (NSCAW) and the National Surveys of Children’s Exposure to Violence but also many under-utilized datasets that have been archived at the National Archive of Criminal Justice Data, National Data Archive on Child Abuse and Neglect, or other data repositories. For instance, Brown and colleagues (2021) conducted a latent class analysis of NSCAW data and, consistent with trauma dose, identified differential effects of child maltreatment and IPV co-occurrence on children’s externalizing behaviors across time and age groups. Lyons et al. (2020) provide an innovative case control study with data from the National Violent Death Reporting System, comparing cases of intimate partner homicides involving a child’s death with live controls.

If you will be collecting new data, better research on trauma interconnections and dosage can be accomplished by simply adding multiple measures of trauma or using measures that already assess trauma dosage, such as the Juvenile Victimization Questionnaire or the ACE-IQ (Hamby et al., 2011; WHO, 2018). Then, be sure to incorporate these measures into your model and connect them to your outcome indicators. The Pu and Rodriguez (2021) article on spillover and crossover effects in this issue is a good example of the kind of work that can be done, both with existing archived data and new data. The paper by Handley et al. (2021) in this issue shows the power of multiple informants. The papers by Free et al. (in press) and Handley et al. also show the value of person-centered analytical approaches.

There is still much to be done to understand the impact of trauma dosage, and more developmentally informed work, such as the papers by Riggs et al. (2021) and Coe et al. (2021), can help us understand the impact and better tailored interventions, such as caregiving and parenting skills interventions targeting adult survivors who were exposed to IPV as children. Their efforts are good examples of other needed research that incorporates contextual information about adversity and trauma. In addition to developmental issues, we need more research on other aspects of dosage, such as number of perpetrators, chronicity, victim-offender relationship, and community context.

Another accessible-yet-important research option is a scoping or systematic review of existing literature. Violence and trauma research is in urgent need of more integrative and synthesizing work. There are good meta-analysis programs, such as Review Manager (https://training.cochrane.org/online-learning/core-software-cochranereviews/revman), that are available for free to promote more quantitative reviews, and guidelines such as those produced by the Cochrane Collaboration to help people rate bias. On the qualitative side, we think the relatively new qualitative interpretive meta-synthesis (QIMS) is a promising approach for aggregating the nuanced insights available in many qualitative works (Storer et al., 2020).

From Documenting Risks and Consequences to Uncovering Pathways to Overcome

One of the broadest trends of the last 20–25 years has been the rise of positive psychology and related work that emphasizes well-being and flourishing, not just eliminating symptoms. This work is increasingly being integrated into research on victimization and trauma, with the rise of concepts such as posttraumatic growth, positive childhood experiences, poly-strengths, and an increase in research on resilience (Bethell et al., 2019; Hamby et al., 2018; Tedeschi & Calhoun, 1996).

We believe that, just as the last 20 years of research on trauma dosage has transformed our understanding of the true burden of trauma, research on interconnections among strengths and the “dose” of one’s overall portfolio of strengths will produce a similar transformation. This work has many parallels to the important work on integration and bridge-building that has happened in other fields. There are rich literatures on many categories of strengths, for example, on emotion regulation and social support, and these can inform work on identifying the set of assets and resources that best prevent and/or overcome trauma (Mariscal, 2020).

We hope to see the field shift even more in this direction in the coming years. Although there is still work to be done on trauma dosage and interconnections among traumas, we need to invest equally in studying the pathways to overcoming trauma. Trauma is far more pervasive than we formerly recognized—with at least 70–80% of the population experiencing at least one victimization, worldwide (Hamby et al., 2021). The coronavirus pandemic has added to everyone’s trauma burden as well, like other non-violent forms of trauma. Although these are grim statistics, what they also mean is that there is substantial untapped wisdom among the many people who have been able to thrive and build good lives despite high levels of trauma (including many professionals in family violence and related fields).

In terms of specific suggestions, one high priority is head-to-head comparisons of different strengths and resources. We urgently need to know which strengths provide “the most bang for buck” in helping people overcome trauma, and the only way to do that is to compare strengths, not just show that any given strength is a generally good thing to have. If you only have 2 or 6 or 12 sessions with a client, would it be more helpful to focus on emotion regulation, social support, or meaning making? It is surprising and unfortunate that we cannot answer such simple questions yet. Again, archived data can be helpful, and meta-analyses and QIMS have potential too. For people who will be collecting new data, there are many brief measures of strengths that can be added to surveys. We especially recommend adding measures of meaning making, an understudied topic in violence research that has shown great promise in our work (Hamby et al., 2020). We also encourage qualitative work to uncover under-appreciated strengths that help people overcome trauma. For all kinds of projects, research designs and analytic strategies should focus on uncovering causal mechanisms when they can.

Paying Attention to Race, Gender, and Other Personal and Community Characteristics

One of the most important and most welcome trends of recent years has been greater attention to individual and group differences that contribute to the social determinants of health (e.g., Walters et al., 2009). Many personal and community characteristics are central features of one’s identity that affect many aspects of life. These include race, ethnicity, gender, sexual orientation, social class, disability status, and other characteristics. Microaggressions and bias-motivated offenses add significantly to many people’s trauma dosage, and we need to better reckon with that in our research (Comas- Díaz et al., 2019; Hamby et al., 2021). Further, we need to study how racial and historical trauma might compound other traumas and victimization experiences among individuals from minoritized groups.

As with other topics, it is easy to say we need more research in these areas, but we would like to highlight a few specific issues. First, we will note that there are entire fields of study that have hardly been touched due to the overwhelming dominance of the mainstream culture. For example, as we far as we are aware, only recently have there been any published empirical studies on whether violence prevention works as well for LGBTQ+ people as for heterosexual, cisgender people (Coker et al., 2020; Waterman et al., 2021). The answer to that question is likely no, which is hardly surprising given that most programs do not address the full spectrum of sexual and gender identities and that violence can manifest differently for different groups. For example, being “outed” without consent can be a unique problem in LGBTQ+ relationships that is not addressed in mainstream programming.

Further, much research on these characteristics has been conducted as an afterthought, an add-on analysis (Banyard & Hamby, in press). Unfortunately, even much of the research on this topic does not unpack the reasons for health disparities, which largely have to do with disproportionate vulnerabilities across groups. No personal characteristic plays a causal role in violence or victimization. For example, people do not offend because they are male, even though men are many times more likely to commit many offenses (such as child sexual abuse and intimate partner homicide). It is because masculinity can be associated with other problems, such as the glorification of male aggression and sexual exploitation in the media, that male gender is associated with elevated risk. Unlike someone’s gender, risk factors like toxic masculinity can be the target of intervention and thus are better research topics than simply exploring gender differences (for example) without further unpacking. Regarding race, the correlated factor that has received the most study is the greater likelihood of poverty, but other factors need more study, including how racism leads to minority stress and microaggressions, exacerbating the trauma dosage for people of color.

In this case, we caution against the common path of conducting post hoc moderator analyses that simply explore group differences in an atheoretical and decontextualized manner (Hamby, 2015; Banyard & Hamby, in press). One option is making use of archived datasets that include multiple forms of violence experienced (and associated risk and protective factors) in specific minoritized communities, such as data on different forms of victimization among Latinx teens and women (Mariscal et al., 2021; Sabina et al., 2014). If you are collecting new data, instead of simply collecting various demographic characteristics, pick the ones you are most interested in and add measures that capture the sociocultural factors that are associated with those characteristics. On the risk factor side, that could include measures of toxic masculinity, microaggressions, minority stress, acculturation, or codes of honor that inhibit helpseeking. On the protective factor side, that could include measures of social support for extended family, cultural identity, resistance strategies, cultural values, faith and spirituality, and practicing cultural traditions (Mariscal et al., in press).

Importantly, qualitative approaches and community-based participatory action research (CBPR) can be especially useful for authentic engagement of minoritized and oppressed communities in deciding what to research and how research could be done. Photovoice, storytelling, and testimonios are creative approaches in which the participant guides the decisions around what and how stories will be told. For authentic engagement, trust-building and recognizing and learning from the participants’ expertise and ways of knowing are critical elements of the process. Authentic engagement also entails involving participants in the design, data collection, analysis, and interpretation, dissemination efforts, and the next steps–key elements of CBPR. This means that research cannot be only for the benefit of the researchers and the field. Authentic engagement research needs to be beneficial to participants and their communities. In a true collaborative spirit, this means that while we learn, gather valuable information, share the findings, and discuss and implement next steps, we build capacity among minoritized communities for them to continue their exploration and transformation of the issues they care about, use the approaches that most resonate with their ways of knowing, and tell the stories they want to tell.

Looking Ahead

There is still much more work to be done in terms of policy, research, and prevention and intervention models for the interconnections among child abuse and domestic violence and other manifestations of trauma dosage. Scientists have an important role to play in helping to reduce the worldwide burden of trauma. There is still much more work to be done to disentangle the web of interconnected family violence and youth poly-victimization and their mechanisms and consequences. Similarly, much more research is necessary to understand different pathways to resilience and posttraumatic growth. Improved understanding of trauma dosage and resilience is critical for developing community-based primary prevention programs as well as trauma-informed intervention programs that adequately and effectively address the needs of families impacted by violence, particularly considering minoritized populations and social determinants of health.