This the story of the Domestic Violence Program Evaluation and Research Collaborative (DVPERC), an ongoing, regional collaboration of domestic violence practitioners, policy-makers, and researchersFootnote 1 from the New England area of the United States (see Table 1 for list of members). Our shared goal is to improve outcomes for survivors and their families by bridging practice and research. Over the years, DVPERC has attended to both process (e.g., authentic, respectful relationships) and outcomes (e.g., rigorous research that is useful to survivors and programs), and we have navigated many challenges along the way.

Table 1 DVPERC member organizations and researchers
Full size table

In this article, we offer an honest description of DVPERC in order to illuminate one form of community based-participatory research (CBPR) for those interested in forming their own practitioner-researcher collaborations in the DV field. We recognize that DVPERC is but one of many practitioner-researcher collaborations that have flourished over the years and been described in the literature. Indeed, we draw upon this rich literature throughout the article to highlight the ways in which DVPERC is both similar to and distinct from other collaborations (see Ragavan et al. 2018 for a review of varying CBPR collaborations and projects).

The article is structured into five sections. The first contains an introduction to the authors of the article and our agencies. The second provides a brief overview of CBPR, including the six values that guide DVPERC. The third details DVPERC’s history, the fourth describes the DVPERC model, and the last section describes the model’s benefits, challenges, and limitations.

Who We Are

Consistent with the tenets of CBPR, especially the principle of transparency, we begin with brief introductions of the six authors of this article in order to help orient the reader to whose perspectives are being shared (and, thus, whose are absent). Together, we comprise two academic researchers (Kristie Thomas and Lisa Goodman) and four administrative-level DV practitioners (Elizabeth Schön Vainer, Deborah Heimel, Ronit Barkai, and Deborah Collins-Gousby). Each of us has been a consistent member of DVPERC since its inception. Although we carry the “institutional memory” of DVPERC, the perspectives we share in this article are ours entirely. We cannot represent nor capture the perspectives and experiences of other members.

Our perspectives are undeniably influenced by our social location. As a group, we all identify as cisgender women, but we vary in our professional backgrounds, race, immigration status, class background, education, sexuality, religion, and survivor status. As a result, we hold unequal levels of privilege and power both within our author group and in DVPERC. First and foremost, the researcher authors hold a great deal of socially conferred status and resources (e.g., job stability, financial resources, time, flexibility) as tenured professors at academic institutions (Simmons College [Kristie] and Boston College [Lisa]). These privileges are intensified by other aspects of our social location, especially our race (i.e., White).

Among the practitioner authors, those of us with fewer intersecting oppressed identities inevitably have more privilege than those of us with multiple oppressed identities. Beyond personal identities, our role as DV program administrators afford us with a certain level of privilege and resources that DVPERC members in frontline positions do not possess. We (all six of the authors) acknowledge that these sources of privilege inevitably influence our interpretation of DVPERC and its work, and we discuss this issue more specifically later in the paper.

As described in Table 2, the DV programs that we (practitioner authors) represent have long histories in the DV field. Three programs are mainstream, and two are culturally-specific in nature.Footnote 2 Although some programs are more well-resourced than others, all are struggling in the current funding climate – a reality that makes participation in DVPERC difficult. We often have to wear many hats in addition to that of administrator; and yet, as we will discuss, participation is even more of a challenge for practitioners from other DV programs.

Table 2 Practitioner Authors’ Roles and Organizations
Full size table

Before moving to the next section, we offer brief personal perspectives on what DVPERC means to us. In our discussions about this article, we felt it important to convey the deep connections we have with one another and other DVPERC members.

To us (the practitioner authors), DVPERC creates an intentional space to reflect and explore the complex and layered nature of the work. The partnership allows for a genuine give and take: through respectful conversation and collaboration, we generate knowledge and evidence-informed practice that is informed by our perspective and expertise. Finally, and perhaps most importantly, the collaborative nature of DVPERC offers a vehicle for elevating survivor voices as we bridge research and practice.

To us (the researcher authors), DVPERC is a community that produces research findings and products that are practical, relevant, and rigorous in the context of mutual learning, support, and friendship. Building authentic relationships with people who work closely with DV survivors and their families is a welcome contrast to the academy, which tends to hover at 30,000 feet. Although DVPERC is a “labor of love” that needs considerable nurturing, it is absolutely worth the effort.

Community Based Participatory Research

From the outset, DVPERC was informed by the tenets of community based participatory research (CBPR). As will be described in detail in this special issue, CBPR is an approach in which researchers and community members (e.g., practitioners, constituents) work to include the voices of all involved (Israel et al. 1998). In CBPR, power and decision-making are shared at every step of the research process, which is made possible by thoughtful and ongoing relationship building (Minkler & Wallerstein, 2010; Yuan et al. 2016). Cultivating strong relationships and sense of community facilitates trust and transparency about intentions, needs, and concerns; all of which are essential elements of successful CBPR projects (Israel et al. 1998; Goodman et al. 2017b; Mouradian et al. 2001; Yuan et al. 2016).

The values and processes of CBPR align well with those of the DV field, which builds on the principles of empowerment, safety, and equality as a counter to the disempowerment, unsafety, and inequality at the core of a violent relationship (Lehrner & Allen, 2009; Cattaneo & Goodman, 2015; Goodman et al. 2017b; Schechter, 1982). In doing so, it engages the DV community in the process, producing research findings that can have a greater and more immediate impact than research that is disconnected from the DV community (National Violence Against Women Prevention Research Center 2001). CBPR is also a powerful mechanism for mitigating community members’ negative experiences with “drive-by researchers,” which often leave them feeling exploited and disempowered (NVAWPRC, 2001).

According to Goodman et al. (2017b), there are six core values that should guide CBPR in the DV field: 1) create relationships of transparency and trust; 2) build on each party’s strengths, resources and interests; 3) strive for equitable decision-making and mutual accountability; 4) foster responsiveness to the ongoing and evolving needs and priorities of all stakeholders; 5) attend to equal distribution of structural and individual power; and 6) facilitate shared ownership of project products. These six values have consistently guided the development and work of DVPERC. In the next section, we describe the early history of DVPERC, organized chronologically and according to how CBPR values informed each stage of the process.

The History of DVPERC

Creating Relationships of Trust and Transparency

The roots of DVPERC extend back to the year 2011, when practitioners from three Massachusetts-based DV agencies began meeting to discuss increasing pressure to evaluate and demonstrate program effectiveness. These practitioners (which included Deborah H. and Ronit) were especially concerned that funders would require them to evaluate outcomes with measures that were neither relevant to survivors’ goals nor feasible for their work. It was a valid concern; at that time, there were very few validated outcome measures that were informed by practitioner and survivor wisdom that could be used in place of funder-prescribed measures (Cattaneo & Goodman, 2015; Goodman et al. 2015a; Sullivan 2011). They decided to develop their own outcome measure, and asked Lisa – a DV researcher and board member at one of the agencies – to assist them. She, in turn, invited Kristie, a DV researcher with prior experience as a frontline DV advocate, to join the group. Although it was not clear initially, DVPERC had been born.

From the outset, initial group members were intentional about infusing CBPR values into the work. A primary example was the commitment to transparency about needs and goals. There were many conversations in which members discussed what they needed to get out of the project in order for it to be worth their time and how they would manage differing needs. Everyone wanted to produce a useful, ethical outcome measure and to do it in a collaborative way that elicited staff buy-in and survivor voice. The group also was unanimous in wanting a measure that would evaluate outcomes that were key to practitioners and survivors and would pave the way for new knowledge about what was working (and not working) in their programs. In addition, the researchers were honest that, given their own academic demands, they needed to develop a measure that was not only reflective of survivors’ own conceptualizations of success but that also would pass scientific scrutiny and be published in a peer-reviewed journal. Practitioners were clear that, although publication was not a primary need for them, they saw the benefit of having a validated measure that they were involved in developing.

The commitment to transparency was in some ways a necessity – no one had time to waste – but it also stemmed from prior research experiences. Several practitioners entered the project having had negative experiences with other researchers, and they wanted to ensure that they would not be taken advantage of again. For example, I (Ronit) had a researcher send me a packet of surveys in the mail with instructions to mail them back when clients had completed them. There was no offer to help or even meet in person to discuss the study. The worst part, though, was that I never heard anything from the person after I mailed the surveys. I had done all of this work and don’t even know what the study found. Conversely, the researchers entered the project having had regrettable experiences doing research that was not CBPR and with a deep commitment to avoid the same mistakes (for examples see Goodman et al. 2017a). As a result, members prioritized having authentic conversations about people’s needs and experiences, which built trust and set the stage for the co-creation of knowledge.

Building on Strengths, Resources and Interests

The initial group quickly realized that the project they envisioned was bigger than the original members could handle and that there was great interest in the project outside the group. Many local DV agencies were having conversations about outcome evaluation, and it seemed essential to include those perspectives and needs. The first step in reaching out to other agencies was for the researcher authors (Lisa and Kristie), to attend one of the monthly regional DV stakeholder meetings to describe the project. The political capital they had built with already participating programs, as well as a professional connection that Kristie had with the person who facilitated those meetings, was helpful in establishing baseline credibility with other programs. Jagosh et al. (2015), call this “trust by proxy” and describe how it can be an important starting point in developing a researcher-practitioner relationship.

The next step involved the researcher authors meeting individually with frontline and administrative staff at interested programs at their offices or via phone. CBPR values informed our approach to these meetings: we focused on getting to know each other’s strengths and resources, discussing program features and needs, and brainstorming ideas. These interactions, combined with the researchers’ subsequent follow-up and follow-through regarding idea integration, helped build upon the nascent level of trust earned initially by proxy. Indeed, interest among other programs grew rapidly; within a few weeks, staff from 15 programs had signed on to participate. We (the practitioner authors) felt that we had finally found researchers who understood the daily realities and challenges of DV work. In turn, we (the researcher authors) were excited about the extensive expertise and enthusiasm practitioners brought to the group and felt confident about their dedication to the project.

Equitable Decision-Making, Power, Mutuality, and Responsiveness to Needs

As our group coalesced, we embarked on our first project – what became the known as the “Empowerment Study,” a three-year project involving surveys of hundreds of survivors across programs that ultimately led to the development of two validated measures: Measure of Victim Empowerment Related to Safety (MOVERS; Goodman et al. 2015a; Goodman et al. 2015b) and the Survivor-Defined Practice Measure (SDPS; Goodman et al. b).

Throughout the study we strove toward consensus-based decision making, mutual accountability, and responsiveness to DVPERC members’ needs and priorities. Holding these values constant helped to facilitate collaboration and minimize disagreement. When people had concerns, we talked things out with one another until everyone felt heard and ready to move forward. As an example, we did not originally intend to develop the Survivor Defined Practice Scale. The idea emerged organically when several practitioners wondered how we would know whether their approach to the work shaped the very success we were working to conceptualize. Subsequent conversation led to the idea of developing a formal measure of survivor-defined practice and validating it in the same study. This, we thought, could be the foundation for exploring whether survivor defined practice is linked to positive survivor outcomes – an important question for programs and building the knowledge base. Although everyone was on board with developing the measure, there was some concern about who would see the data, especially if the results were not positive. We talked at length about these concerns and collectively developed a plan for analyzing and reporting data that was satisfactory to everyone.

The process to develop the measures was time consuming and labor intensive. To develop the constructs, the researchers conducted a scholarly literature review and held focus groups with survivors in member programs, trying to learn about how they thought about successful outcomes. We brought our findings into monthly DVPERC meetings where we discussed them in the context of practitioners’ on-the-ground experience and understanding of success, eventually developing the main constructs of interest, safety-related empowerment and survivor-defined practice (see Goodman et al. 2015a for a detailed discussion of this process).

Next, we spent several months co-creating items for the two planned measures and discussing what additional information we wanted to gather during the planned data collection process. Although the process of building the survey packet illuminated differences among some of the programs in terms of focus and priorities, our strong relationships and shared goal of creating useful measures allowed us to achieve a consensus that reflected both the state of the research and the experience of survivors and practitioners.

Once the survey packet was finished, we administered it to more than 300 survivors across participating programs, which included shelters, transitional living programs, community-based programs, and hospital programs. We collectively decided that program staff would not administer the surveys. It was important to minimize the potential that survivors would feel obligated to participate and to maximize their honest participation, especially when responding to questions about the use of survivor-defined practice. In addition, programs worried that the involvement in survey administration would be an undue burden on staff. Instead, the first two authors and a team of trained graduate students from their academic institutions administered surveys in each program. The surveys had been developed in English and translated into Spanish; we provided translators for the few participants who spoke other languages. Programs supported these activities by offering childcare during survey administrations, offering emotional support to participants when necessary, and communicating with the group regarding concerns that emerged (See Goodman, et al. 2015a for further description of the process).

Sharing Ownership of Project Products

Once data were collected, the researchers commenced with data analysis to develop the final validated measures. Practitioners engaged in this process to the extent that they wanted to (e.g., looking at analyzed data to determine the factor structure of the measures, reviewing the final phrasing of items to ensure they would make sense to survivors). Once the measures were developed, the researchers and Deborah H. (who was most active in working with the data) brought those findings to the rest of DVPERC for their feedback. We also discussed how we would disseminate the measures to the public and the pragmatics of how the measures could be used effectively in programs.

The dissemination process was quite involved and led to several products. As mentioned, researchers were interested in writing up the process and measures in peer-reviewed academic articles. Only one of the practitioners (Deborah H.) was interested in co-authorship and joined the researchers in that venture. Instead, practitioners were very keen on ensuring that the measures would be maximally relevant for their programs. For example, although the safety-related empowerment measure was intended to be an outcome measure, practitioners initiated discussions about how it could be used as a practice measure as well, to guide work with individual survivors. They also wanted to ensure that programs would have the knowledge to actually use the measures in their own contexts. This led Lisa and Kristie to visit all interested programs to work with staff on developing procedures for using the measures. They also worked with Deborah H. to create a user-friendly online guide describing in detail how to administer, score, and interpret the MOVERS measure (Goodman et al. b). Finally, practitioners wanted to use the information gathered during data collection to improve their programs, so they asked Lisa and Kristie to prepare reports for each program, which they did.

DVPERC Evolves into an Ongoing Collaborative

The Empowerment Study was energizing, and, as relationships developed among collaborators, we grew excited about the possibility of expanding our impact by working on new projects. By this point, we had begun to refer to ourselves as the DVPERC. Having a name helped to legitimize the project and made it easier to invite other programs to join us. A collective sense arose that DVPERC should not end when the initial project did. For us (practitioner authors), the expansion and formalizing of DVPERC was helpful because it allowed us to deepen our connections and support each other, particularly with program evaluation efforts. Having designated meetings provided time to discuss best practices, brainstorm new ideas and strategies, and troubleshoot situation-specific challenges. For us (researcher authors), it was thrilling to imagine a vehicle for co-creating knowledge with practitioners on an ongoing basis. Prior to DVPERC, neither of us had experienced an ongoing CBPR partnership, and frankly, we did not know it was even a possibility. We were excited for DVPERC to evolve.

We embarked on two new research projects that took us in different directions and led to many smaller meetings in addition to the larger meetings with all DVPERC members. Kristie and three other DVPERC programs began a project to explore the needs of homeless survivors placed in Massachusetts emergency assistance hotels, a major crisis and concern for programs at that time (see Thomas & So, 2016). Many other programs joined with Lisa to develop and validate what would become the Trauma Informed Practice (TIP) Scales (see Goodman et al. 2016a; Sullivan and Goodman 2015). This project emerged as a response to the growing tidal wave of interest in trauma-informed care among DV programs and researchers nationally (e.g., Phillips et al. 2015) and locally (see Lewis-O’Conner & Chadwick, 2015 for TIP among one of the DVPERC member programs). As with the Empowerment Study, programs were eager for tools to help them assess the extent to which their services aligned with the tenets of trauma-informed practice, and they wanted to be involved in the process. Because the TIP Scales project involved other researchers working with national DV resource centers (e.g., Cris Sullivan, Julia Perilla, and Josephine Serrata) knowledge of DVPERC spread to a larger audience and set the stage for later projects with these and other collaborators (e.g., an online toolkit for researchers on how to conduct CBPR with DV communities; Goodman et al. 2017b).

These projects were ambitious and consuming. We decided to pause before embarking on any new projects in order to focus on other tasks, replenish our energy reserves, and forget the pains of primary data collection. Moreover, membership was in flux, which was not surprising given that several years had passed since DVPERC began. New programs and researchers were joining and some existing members were taking a break. Even among stalwart programs, staff turnover meant new representatives who were unfamiliar with the origin of DVPERC. As others have noted, staff turnover is a considerable threat to the stability and success of collaboration (Macy and Goodbourn 2012; Sullivan et al. 2017). As we will describe later in the article, it has contributed to moments of mission drift over the years regarding the “who” and “what” of DVPERC. In the next section, we describe the model that emerged out of this transition time, and which is current as of the time of this writing.

The DVPERC Model

Within the larger frame of CBPR, the current DVPERC model can be characterized by five elements: an informal structure, varied and fluid membership, regional in scope, ongoing format, and unfunded (see Fig. 1). Although these elements are not unique to DVPERC, they coalesce to form a model that is distinctive and, to our knowledge, rare. This section provides a brief description of the elements, followed by a section on the model’s benefits and limitations.

Fig. 1
figure 1

The Domestic Violence Program Evaluation and Research Collaborative (DVPERC) Model. Note. CBPR = Community-Based Participatory Research

Full size image

Element #1: The DVPERC Structure is Informal and Flexible

Ultimately, we decided to adopt a relatively informal and flexible structure for DVPERC. Although we are united by a desire to bridge DV research and practice and a commitment to mutuality and co-learning, our structure is flexible: Following the first year of monthly meetings, we moved to a bimonthly structure. Each two-hour meeting takes place in a conference room at Jewish Family & Children’s Service. Meeting topics are co-decided and vary based on interest and need. For example, in addition to discussing our ongoing research, we have hosted CBPR teams from other states to learn about their DV work, heard from practitioner members about their experiences at national conferences, discussed current debates in the DV field (e.g., victim/perpetrator dichotomy) met with national agencies to discuss DV-related policies at the state and federal levels, and served as a sounding board for a practitioner writing a position paper on fathers and child welfare. Kristie, Lisa, and Deborah H. share primary responsibility for communicating with members about meetings; this work involves soliciting ideas for meeting topics and maintaining an ever-expanding list of names and contact information.

Although our regular meetings are the bedrock of DVPERC, some of the research and evaluation work happens in meetings with smaller configurations of members. Once programs and researchers decide to conduct a specific research project, for example, we hold separate meetings at the site of the program(s) involved. Or when programs want to discuss implementing the results of a specific study, researchers may meet with them individually. For instance, several programs have wanted more individualized discussions about the challenges they face in using measures to conduct evaluations (e.g., when and how often to administer, how to systematize data collection procedures, etc.) We (researcher authors) have had many meetings and phone conversations with individual member programs to discuss their evaluation efforts. These conversations are always mutually beneficial in terms of learning and idea generation.

Element #2: DVPERC Membership is Varied and Fluid

To become a DVPERC member, one need only be a practitioner or researcher interested in bridging DV practice and research. Members can come once or every time, as they are able and interested. This low threshold, “open door” policy is in response to the demanding and unpredictable nature of DV direct service work. It simply is not realistic for staff from each member program to attend every meeting. In some cases, staff at particular programs have to pause participation indefinitely because of organizational-level issues, but they are welcome to return at any time. As others have noted, participation in collaborations is inherently fluid and there is often ambiguity regarding who exactly comprises the community (Yoshihama & Carr, 2002). This situation certainly applies to DVPERC. There are usually anywhere from 15 to 25 attendees. Initially, there were substantially more practitioners than researchers in attendance, but that has shifted over the years. Although practitioners still outnumber researchers, the ratio is slightly more balanced currently.

The programs involved in DVPERC are primarily mainstream in nature (vs. culturally-specific), but vary in terms of size, scope, and services. Some are distinct agencies; others are smaller programs housed within hospitals, service organizations, and state agencies. Many of the programs offer residential services in the form of emergency shelter and transitional housing, and they serve urban and suburban communities. DVPERC practitioner members include administrative as well as frontline staff; however, it is more challenging for frontline staff to commit to regular attendance given the crisis nature of their work. As a result, they are more likely to attend meetings or be involved in projects when the topic is of particular relevance to their specific work and population. DVPERC members in administrative roles try to support frontline staff participation by offering to carpool, ensuring that responsibilities are covered, and keeping them abreast of discussions at meetings they missed. Nevertheless, as we will discuss later, frontline staff participation is just one of several membership challenges for DVPERC.

Researcher members vary in terms of discipline and violence-related research focus. They span counseling psychology, community psychology, medicine, and social work, with areas of focus that include (but are not limited to) poverty and homelessness, trauma-informed practice, community-based interventions, children who experience DV, and the criminal legal system. In addition, career stage varies from doctoral student to full professor.

Although this was not an easy decision, DVPERC was developed as a collaboration between researchers and practitioners, many of whom are themselves survivors, but not program clients. Early participants wanted to provide a space for members – especially practitioners – to share their doubts, confusions, and challenges in ways that they thought would not be useful for their clients, and perhaps even harmful. Survivor voices do shape our project, not only because quite a few of us identify as survivors and carry those perspectives with us, but because all of our projects include qualitative focus groups with program participants. Still, as we will discuss, there are substantial drawbacks to this decision.

Element #3: DVPERC is Regional in Scope

DVPERC members are mainly from New England, with the vast majority located in Massachusetts. The regional emphasis is an intentional strategy to ensure face-to-face interaction. We select dates and times that work best for the majority of us and meet at a site that has ample parking and is close to a major highway. We have discussed rotating meeting location, but have done so only several times because very few programs have the capacity to host large groups. We also have discussed the option of virtual attendance; up to this point, however, we have not had the capacity or the skill to do so. An early attempt at having members call in was unsuccessful and frustrating. As our access to technology increases, we plan to explore how to integrate virtual attendance in the best way possible.

Element #4: DVPERC is an Ongoing Collaboration

DVPERC members made a conscious decision to move from what began as a project-based collaboration to an ongoing one. Unlike project-based collaborations, our work is not guided by deadlines or specific objectives that we must achieve. We will continue to exist as long as there is interest to do so. Having an informal structure and fluid membership has played a role in our longevity. There are enough researcher and practitioner members to support multiple projects at any given time; such involvement sustains the group identity and provides room for members to take a break from research. As one example, Megan Bair-Merritt and Maya Ragavan, pediatricians whose research focus on children who experience DV, joined DVPERC in 2015 with funding to develop a parenting-focused phone app for survivors. Several members were interested in being involved and others were not. Being ongoing meant that we continued to hold bimonthly meetings as a group even as a small number of practitioners and researchers met separately to work on a particular project.

Element #5: DVPERC is Primarily Unfunded

We believe that one of the most surprising aspects of DVPERC is that it operates without any external funding. Over the years, we have secured small amounts of money from various sources to help allay some of the costs of specific projects, but even in those cases, volunteer labor was essential. For example, during the Empowerment Study, although we had help from volunteer graduate students, the researchers administered the majority of the surveys, travelling to shelters and agency offices across Massachusetts and other states. At the same time, the practitioners organized scheduling and other logistics on top of all of their other job responsibilities. As we will discuss in the next section, there are unexpected benefits to being unfunded; however, the downsides tend to outweigh those benefits.

Benefits and Limitations of the DVPERC Model

Benefits of the DVPERC Model

Authentic relationships

Perhaps the most important benefit of the DVPERC model is that it facilitates authentic relationships rooted in trust and mutuality. As many others have noted, trust is essential for any successful practitioner-research collaboration, but it is especially in the DV field given its history of researcher harms and subsequent practitioner distrust (Edleson and Bible 2001; Davidson and Bowen 2011; Goodman et al. 2017b; Hamberger and Ambuel 2000; Richie 2012; Schechter, 2017).

Being a regional collaboration contributes to authentic relationships in two ways. First, it allows for regular, face-to-face contact, which promotes and accelerates relationship building. By gathering in person in our large group meetings, we establish a basic familiarity with one another that opens the door to assemble in smaller configurations, not only for work-related issues but also social interests. For instance, we are just as likely to meet up for evaluation-related meetings as we are to catch up over a meal or cup of coffee. Proximity also means that we can attend each other’s work events (e.g., agency fundraisers, panel discussions, trainings). Such regular, face-to-face interaction has facilitated trust and friendship. Indeed, the CBPR literature is replete with examples of how important it is for collaborators to participate in each other’s work (e.g., Bloom et al. 2009; Edleson and Bible 2001). And, as others have noted, these interactions also inform our research questions and projects (Hamberger and Ambuel 2000).

Second, being a regional collaboration means that we have at least a working knowledge of the institutional systems and major players that affect one another’s work. Such familiarity allows members to support one another in ways that extend beyond collaborative research. As one example, when practitioner members in Connecticut were in need of shelter for a client and her children, they used the DVPERC network to connect her with several of the Boston-based practitioners. As another example, Elizabeth and Kristie have developed a mutually beneficial tradition: when the Simmons School of Social Work contacts Elizabeth each spring about supervising an MSW intern for the following year, Elizabeth asks Kristie to identify strong candidates who would be a good fit. For me (Elizabeth), it is a relief to not have to stress about whether our program and clients will get the best possible intern. For me (Kristie), I appreciate knowing my students will be at a great program working on an issue they are passionate about.

Being an ongoing collaborative is another important factor in cultivating authentic relationships. With longevity comes abundant opportunities for DVPERC members to get to know each other well. This is important given the fluidity of our membership. As others have noted, sporadic participation lengthens the time it takes to build trust (see Gilfus et al. 1999). Regarding researcher members in particular, long-term participation demonstrates that they are in it for the “long haul.” This foundation of trust enables participants to take risks with each other. As one practitioner said at a meeting, “There are no sacred cows or bulls; we can be honest and know that we won’t be shut down.” That said, we are cognizant that trust fluctuates over time and circumstance; thus, sustaining it is a continuous process (Jagosh et al. 2015; Maciak et al. 1999).

Finally, the fluid membership style minimizes the chance that members experience DVPERC as an obligation or a burden. Instead, people attend because they want to. This is especially important to us (practitioner authors). We appreciate that we can take a short or extended absence from DVPERC and return to be greeted with a warm welcome and genuine joy. This response makes it easy to reconnect and engage with the group, regardless of whatever time has passed. Ultimately, although people join DVPERC because they are committed to DV work, it is the relationships with one another and the learning that it enables that sustains our group. DVPERC provides a blend of collegial support and professional development, both of which are necessary to mitigate the vicarious trauma and compassion fatigue that so often accompany DV work (Slattery and Goodman 2009).

High relevance

Another primary benefit of the DVPERC model is that its flexibility increases the relevance of its work. Being ongoing means that we can be responsive to new ideas as they emerge. For example, two of Lisa’s doctoral students, Jennifer Fauci and Joshua Wilson, developed dissertation studies based on DVPERC conversations about the current challenges programs face. Anjali Fulambarker, an assistant professor at Simmons College, is working with several DVPERC members and their collaborators to evaluate a trauma-informed policing project. These examples also highlight how longevity allows DVPERC to ensure that CBPR continues through newer generations of researchers, who often struggle with being the “new kid on the block” among the DV community (Hamberger and Ambuel 2000, p. 258).

Fluid membership means a constant influx of new ideas and perspectives. For example, after a Connecticut DV program joined, we had several useful discussions about how CT and MA differ regarding practice, policy, and system response. Similarly, recent discussions about survivor-mothers’ experience of seeking help from advocates who are also mandated reporters began when a program that collaborates extensively with the Department of Children and Families (DCF) joined DVPERC.

Being an unfunded collaborative affords a certain level of freedom to choose the research questions we think are most important. We do not worry about pleasing funders or meeting their requirements; instead, we focus on what is important and interesting to members. For us, (practitioner authors) it is a particularly welcome change to not be constrained by the many agencies and organizations that keep our programs afloat. We can focus on the topics that we want to explore, and we have the flexibility to dive in.

Although we have never formally evaluated the impact of DVPERC, there are many indications that our work is making a difference. The measures we created (e.g., MOVERS, the Survivor Defined Practice Scale, the Trauma Informed Practice Scales) have been adopted by practitioners and researchers from across the country and even internationally. The consistent volume of requests we receive speaks to the hunger that people have for rigorous measures informed by practice. We (practitioner authors) use these measures to guide direct practice, inform program planning, and secure funding. Other products, such as the parenting app described earlier, are directly beneficial to our clients and their families.

Limitations of the DVPERC Model

Limited diversity and inclusion

DVPERC suffers from an overall lack of diversity among members. Most strikingly, very few of the agencies and programs are culturally-specific. One reason stems from DVPERC being unfunded: We are located in a major metropolitan area; travelling even short distances - by car and public transportation - can be time-intensive. It is not uncommon for members to miss meetings or other gatherings because they cannot spare the travel time. This situation is particularly acute for smaller, less resourced programs such as culturally-specific programs, which cannot afford to lose a staff member for a stretch of several hours. Indeed, the primary reason that Journey to Safety is able to participate, despite their small size and budget, is that they host the large meetings at their space, making it easier for them to attend. Although not a panacea, funding would allow for programs to be compensated for their time and effort, a strategy that has been helpful in other CBPR projects (e.g., Sullivan et al. 2005).

Beyond funding, however, there are other possible reasons why culturally-specific programs are underrepresented. For example, they might have had difficult experiences with the mainstream programs that attend, or they have felt excluded in some systematic way from DVPERC based on the topics we choose or the knowledge we have produced. As we will describe later, further conversation with them as part of a planned outreach effort will enable us to learn more about the barriers that prevent their participation.

In addition, practitioners in administrative roles are overrepresented in DVPERC. Practitioners who act as frontline staff at DV agencies have less autonomy in their schedules compared to administrative staff, making it more difficult for them to attend meetings. Further, although DV programs generally are not able to pay any employee well, frontline staff typically earn the lowest salaries, which can mean not having a car or other resources necessary to get themselves to DVPERC easily.

The overrepresentation of practitioner members in administrative versus direct service roles also has implications for DVPERC’s racial and ethnic diversity. A recent survey conducted by the Massachusetts Women of Color Network found that the large majority of leadership roles in domestic and sexual violence organizations in Massachusetts are held by White women (Prabhu 2017). For example, Women of Color held only 12% of the Executive/Co-Executive Director positions and only 37% of the supervisor/manager positions – but they held 55% of the advocate positions. Evidence suggests that this racialized imbalance of power is not limited to Massachusetts. A survey of DV programs in North Carolina also found that White people were disproportionally represented in leadership roles in DV programs, while people of color were concentrated in frontline, direct practice roles (Macy, Giattina, Parish, & Crosby, 2009).

These limits on DVPERC’s diversity has serious consequences for the kinds of knowledge it produces. The research we conduct reflects the questions and interests of the people in the room. In turn, research affects policy and practice. From the start, the DV movement has been marked by the marginalization of voices of oppressed communities (Kim 2013; Mehrotra et al. 2016; Richie 2012) and the adoption of a one-size-fits-all approach that does not center the unique needs of marginalized populations (Kim 2013; Mehrotra et al. 2016; Richie 2012). Culturally-specific programs emerged to fill this void; however, like the communities they serve, they operate on the margins with more complex crises, fewer resources, a lack of public support, and even outright violence directed at them.

Given these many challenges, it is much more difficult for culturally specific programs to participate in collaborations such as DVPERC. In fact, for all the reasons we have described, the current DVPERC structure may prohibit participation of these most marginalized programs. Although we describe ourselves as inclusive and welcoming, we have not engaged in extensive outreach to attract new members. A one-time invitation to join might suffice for mainstream programs; however, culturally-specific programs might require more relationship building to overcome their justifiable hesitations (e.g., historically-rooted fears of research). As a result, we are not hearing the questions that these programs are struggling with, engaging them in the co-creation of knowledge, or producing findings that would be useful to their clients. The absence of culturally-specific programs, the dearth of women of color in leadership positions, and the decision to exclude current clients from membership coalesce to mean that the DVPERC model plays a role in perpetuating the tradition of marginalizing already marginalized communities.

Although there are no easy solutions for addressing these problems, DVPERC members have been, and continue to be, committed to finding them. We are currently developing a needs assessment that will explore, among other topics, member and non-member perspectives on barriers to DVPERC involvement and ideas for addressing them. In the meantime, we started a new recruitment strategy that draws upon existing members’ untapped networks. Each member is responsible for reaching out to at least one person that they have a connection with at a non-member program or an inactive member program, especially culturally-specific and under-resourced programs. As one example, the DV agency where Ronit works recently hosted a panel that featured the work of two local culturally-specific programs (one of which attends DVPERC sporadically). Planning that event strengthened her relationship with the panelists and opened the door to a conversation about DVPERC. In addition, two of DVPERC researcher members (Kristie and Anjali) attended the event to show support, meet the panelists, and learn more about their agencies.

Finally, many of our members are part of an initiative started by Jane Doe, Inc. (the Massachusetts Domestic and Sexual Violence Coalition) designed to facilitate deep reflection and honest conversations about the history of racism and marginalization within the DV movement broadly, and in Massachusetts specifically. According to the Massachusetts Women of Color Network, “women of color voices and leadership has been pushed to the margins of

the very work they began” (Prabhu 2017, p. 4). In their model, programs can increase access for women of color by providing opportunities for paid internships, professional development, and mentoring. DVPERC members who are involved with the initiative of Jane Doe and the Women of Color Network will continue to use what they have learned to shape the work of DVPERC.

This larger conversation reflects the reality that DVPERC is not alone in struggling with diversity and inclusion. As others have noted, it is challenging to fully realize the values that guide CBPR due to the oppressive structural forces that hinder participation from socially and financially marginalized communities (Goodman et al. b; Jagosh et al. 2015; Ragavan et al. 2018, Yoshihama et al. 2012). It is incumbent upon all CBPR collaborations to develop creative strategies that prioritize the voices of those with the most at stake; otherwise, they reify the very imbalance that makes CBPR necessary.

Decision-making not shared at every level

A critical component of any CBPR approach is ensuring equitable power distribution among all involved. How to best do that, though, has been a challenge for DVPERC. As we have discussed, there are unequal levels of power and privilege in our group. We (researcher authors) have jobs that afford us with more time, flexibility, and autonomy to do the work of DVPERC (e.g., identifying speakers, onboarding new members, organizing meetings, doing the everyday work of research), resources that our practitioner partners do not have. And yet, shouldering more of the work inevitably made us the “face” of DVPERC and the unofficial “leaders.” Given the power dynamics that exist due to our credentials, we did not want to perpetuate a tradition of researchers doing something to the group versus doing with the group (Goodman et al. 2017a, b; Paradiso de Sayu and Chanmugam 2015).

Several years ago, we brought our concerns to the group to brainstorm how to share leadership more widely without burdening anyone. (We acknowledge that the very act of us bringing it to the group represents a power imbalance.) What emerged was the sentiment that members do not need or want to be involved in every decision, and that they preferred for us to “lead” DVPERC so they wouldn’t have to. For example, they reported appreciating that we would brainstorm potential meeting topics with a few of the more involved members before running ideas by the group. As a result of that conversation, we regularly circle back to issues of power and decision-making. In fact, the very writing of this article led to a helpful reframing of the role we have played since DVPERC began: Elizabeth suggested that practitioners see us as “active facilitators” rather than leaders.

We (practitioner authors) agree that, from the outside, it looks like power is skewed or unbalanced; however, DVPERC simply would not exist without Lisa and Kristie taking the lead on coordination and facilitation. None of us have the time to play that role, and we are glad they are willing. We are happy for them to take the lead role, with consultation, in identifying good topics for discussion and know that when we propose ideas, they are both respected and realized. The exciting aspect of sharing our ideas and wishes with researchers is that they are often able to draw upon their research networks to invite speakers who expand our capacity to address whatever need we have identified. Nevertheless, the process of “checking in” is essential for maintaining the foundation of trust and cultivating group norms of transparency and equality. This, in turn, promotes the opportunity to transfer responsibility as needed.

These struggles highlight how the CBPR values of power sharing and equal decision-making may be operationalized quite differently within and across collaborations, depending on the wishes of group members. In our case, shared power does not equal sharing every task and responsibility. That does not mean, however, that power is balanced between researcher members and practitioner members. On the contrary, the DVPERC format relies on, and therefore reifies, a system that privileges the voices of academics over practitioners. Because we (researcher authors) have more resources, we ultimately have more control, play a bigger role, and garner more of the credit, thereby reinforcing our status as “expert.” As an example, we (practitioner authors) would like to present with the researchers at academic conferences more often, but it is difficult for us to do so. Unlike researchers, most practitioners do not have the flexibility or travel stipends to attend conferences – a harsh reflection of the resource inequalities that exist between academic institutions and DV programs (and other social service agencies).

Mission drift

Although evolution is inevitable, being an ongoing collaboration poses challenges that time-limited collaborations might not experience. In its nearly seven years of existence, DVPERC has experienced continual change in terms of membership, focus, and surrounding context. Despite not having a formal mission, this evolution has led to instances of mission drift. As one example, one of our meetings occurred soon after the 2016 presidential election. There was a sense of urgency that we should act, but no one really knew what that action should be. Additional discussions at subsequent meetings unearthed concern over introducing legislative advocacy as an element of DVPERC as this felt exhausting and potentially redundant with the work of other organizations. Ultimately, we decided that DVPERC was not a feasible or appropriate forum for organized political action. We are much better positioned to serve as a venue for discussing members’ own emotional needs regarding the ripple effects of the current administration, supporting agencies’ individual initiatives, and brainstorming research projects that might bolster those initiatives. As others have noted, it is not uncommon for CBPR collaborations to struggle with whether and how to balance research with a desire for more immediate action and activism (Burke et al. 2013; Maciak et al. 1999).

Labor-intensive

DVPERC requires a substantial amount of time and energy. Having a fluid membership would be difficult for any collaboration; for DVPERC, the challenges are compounded by being rooted in a CBPR approach that prioritizes authentic, intentional relationship building. It is important that we spend time “onboarding” new members (e.g., discussing the history and goals of DVPERC, what involvement entails, and whether it is a good fit for them), which amounts to many meetings, emails, and phone calls. In addition, our low-threshold membership policy means we have to play detective sometimes to assess why people are no longer attending. Is it that they are taking a break but still want to receive emails, no longer interested, or no longer working there? In the latter two cases, we try to contact other people at the program so that the thread is not lost, but it takes time, energy, and thoughtful follow-up.

As an ongoing collaboration, there is a need to fill in the time between projects with activities that keep people feeling connected to DVPERC and each other. It can be challenging to maintain interest without a specific shared project or consistent attendance, but mandating attendance is antithetical to our CBPR approach. We (researcher authors) worry about being overbearing in our efforts to promote attendance, which elicits one of the “check-ins” described earlier. To some extent, our discomfort with not having outcomes fuels some of our worry. Over time, we have come to heed the words of Ann Fleck-Henderson, a member of a similar research-practice partnership in Massachusetts that predates DVPERC: “Process is the product when it comes to collaboration” (as quoted in Gilfus et al. 1999, p. 1205).

Conclusion

DVPERC has been a richly rewarding experience that has led to enduring relationships and the co-creation of knowledge for the field. We hope that we have not sugarcoated the arduousness of the process, nor the challenges we have encountered. The DVPERC model is not perfect; it has serious flaws that, although unintentional, can perpetuate structural oppression in the DV field. That is inevitable (Goodman et al. 2017b). At the same time, members work hard to note and acknowledge such dynamics and to work to undo them to the extent possible. DVPERC is a work-in-progress; being an ongoing collaboration provides the opportunity for continual reflection and improvement. It is also a deeply fruitful and joy-filled process. We hope that this article informs and inspires others to launch their own CBPR collaborations.