Intimate partner violence (IPV) is a significant social and public health problem in the United States. IPV’s impact on long-term negative health consequence has been well-documented, and this increased health burden translates into higher health services use and costs for victims of IPV (Bonomi et al. 2009; Koss et al., 1991; National Center for Injury Prevention and Control, 2003; Rivara et al., 2007; Snow Jones et al., 2006; Ulrich et al., 2003). A 2010 national survey estimated almost 7 million women and 5 million men experience physical or sexual violence or stalking by an intimate partner each year (Centers for Disease Control and Prevention 2010). A recent national survey noted that one in five men admitted to having “pushed, grabbed, or shoved; thrown something at; slapped or hit; kicked, bit, or hit with a fist; beat up; choked; burned or scalded; or threatened with a knife or gun” their intimate partner (Danis, 2003, p.237).

Research conducted on IPV has prioritized ways to improve and provide services and support for victims and their families. Relatively little focus has been given to addressing prevention and intervention with IPV perpetrators. Group intervention approaches, commonly referred to as batterer intervention programs (BIPs), have become the predominant form of treatment for perpetrators of intimate partner violence (Tolman, 2001). The popularity of these groups has increased, and, to date, 45 states have developed formal standards of care for BIPs. These standards, however, are based primarily on policy makers’ beliefs about what constitutes a good program, despite there being little consensus about key intervention components, such as length of program and credentials of program facilitators (Austin et al., 1999; Maiuro & Eberle, 2008; Price & Rosenbaum, 2009).

Most of what is contained in the literature are general descriptions of the theoretical models and approaches for various BIPs (Dutton et al., 2007; Morrel et al., 2003; Saunders, 1996; Saunders, 2008; The Duluth Model, 2011), or experimental studies that test the efficacy of BIPs through measures of recidivism and/or re-assault as outcomes (Alexander et al., 2010; Babcock et al., 2004; Brannen et al., 1996; Dunford, 2000; Feder et al., 2005; Gondolf, 2009; MacLeod et al., 2009; Mills et al., 2012; Morrel et al., 2003; O’Leary et al., 1999; Rempel et al., 2008). This literature has focused primarily on programs that are described as using feminist-psychoeducational or cognitive behavioral change approaches. While there has been some detailed information regarding the impact of such groups on client attitudes, beliefs, and behaviors (Aguirre et al., 2011; Buchbinder et al., 2008; Chovanec, 2012; Pandya & Gingerich, 2002; Rosenberg, 2003; Scott & Wolfe, 2000; Silvergleid & Mankowski, 2006; Wangsgaard, 2000), there is very limited literature from the perspective of professionals who do this work (e.g. BIP facilitators) or whose work relies on BIPs (e.g. law enforcement, victims’ advocacy) and how they perceive these interventions (Pandya, 2009; Silvergleid & Mankowski, 2006). In order to better understand what impact BIPs might have on IPV prevention and intervention, and appropriately inform state standards, more information is needed from those individuals working “on the ground” on what they believe makes good BIPs. This is the first study of its kind to describe the key components of a successful BIP from the perspectives of professionals whose daily work involves interactions with IPV and BIPs.

Methods

Overview

We conducted a two-year ethnographic study of batterer intervention programs for male perpetrators of intimate partner violence against female partners in an urban area in the United States. The aims of the parent study were to describe the content, processes and strategies used in BIPs, and to broadly explore the BIP experience from the perspective of clients, client partners, and professionals who are employed by, or whose work directly relies on, BIPs. An ethnographic approach was taken because such studies allow investigators to interact with participants in their real-life environments and are better suited for obtaining a more in-depth, nuanced understanding of contexts (Giacomini et al., 2000; Patton, 2002). The parent study, therefore, utilized multiple methods of data collection (e.g. participant observation, debriefings and interviews) to broadly explore and investigate the BIP experience. This particular analysis examines what professionals, working in or with BIPs, believe to be the key components of a good batterer intervention program.

Setting

Two batterer intervention programs were selected for participation in this study. The BIPs represented the largest by volume of clients in the area with each servicing 100 or more clients per year. They were also the most experienced programs in the area, having over 10 years of service provision, and at least 5 locations at the time of data collection. Each program used an adaptation of a national BIP model (Duluth and Emerge, 2011) as the basis for its curriculum. In order to ensure participation, and facilitate openness and honesty in our interviews, study participants were recruited with assurances that their statements would be anonymous. The BIP/IPV community and auxiliary agencies (e.g. the judicial system) that work with BIPs is relatively small in the region where the study took place; thus, if detailed demographics were provided on agency roles, length of years in service or otherwise were provided, it could easily identify our participants and represent a breach of confidentiality. We, therefore, grouped our participants into three categories based on their professional domain. These include BIP experts/IPV victims’ advocates, judicial and legal professionals, and policy/human services professionals.

Data Collection

Data collection for this study took place from 2013 to 2015. Recruitment took a snowball sampling approach with two initial study participants from the collaborating BIPs suggesting and facilitating contact with additional individuals they felt would have a valuable perspective to share regarding BIPs. Individuals were eligible for study participation if they had worked directly with a BIP, or worked in an area related to batterer intervention services. This resulted in representatives from a broad range of professions, including BIP facilitators and administrators, IPV victims’ advocates, legal and judicial officials, attorneys, health department employees, and county and state policy makers. The semi-structured interview guide broadly explored the definitions and perceptions of best practices for BIPs, barriers/challenges to intervention, and beliefs about what factors contribute to perpetration of IPV. Questions relevant to this analysis included: 1) What does an ideal batterer intervention program look like? 2) At minimum, what should a batterer intervention program include? 3) What are realistic expectations for a batterer intervention program? All interviews were conducted by a PhD trained anthropologist, in a private space selected by the participant. Each interview lasted between 45 and 90 min. Verbal consent for was obtained prior to each interview. The institutional review board at the University of Pittsburgh approved this study.

Analysis

All interview data were audio-recorded, transcribed verbatim by a trained transcriptionist, and then entered into Atlas.ti, a qualitative data management program, for organization and analysis. We utilized a two-coder iterative approach and member checks to analyze the data. Analysis focused on content and global coding of broad thematic categories and subcategories. First, the first author and a qualitatively trained research assistant each independently reviewed the transcripts line-by-line to identify preliminary themes. Preliminary themes were then reviewed with participant stakeholders for input and feedback. The coders then met to refine each theme and organize each in hierarchical categories. The codebook was once more reviewed by participant stakeholders to ensure reliability and validity and definitions were delineated for each major and minor theme to assist in the final coding step. Each coder then independently re-coded the transcripts using the codebook and met once more to reconcile any differences if needed. Reported themes are those that arose consistently across participants.

Results

A total of 36 semi-structured, open-ended interviews were conducted with professionals who work directly with, or whose work relies on, batterer intervention programs. Our sample included 52 % BIP experts/IPV advocates, 33 % judicial and legal officials, and 15 % policy makers and human services professionals. A majority were Caucasian (75 %) and female (70 %). All had at minimum 5 years of professional experience in BIPs or IPV. Analysis yielded three main thematic categories related to what participants viewed as the fundamental components of a quality batterer intervention program: (1) practices related to optimal BIP structure, (2) desired characteristics for facilitation, and (3) program approaches needed at minimum for promoting change.

Optimal Structure

Participants described what they considered to be optimal BIP structure in terms of individual group size and length of programs for maximizing the potential for impact.

Group Size

While beliefs about the exact number of clients that should be allowed in group varied slightly, there did seem to be universal agreement that BIPs should be limited in size and that bigger groups were less effective. As one participant stated, “I’ve heard in some groups they have 26. How can you do any work with 26? With 26 people, how are you going to learn?” Another similarly described the difficulty of having over 20 clients in group, “When I had 20+ it would take me 20-30 minutes to sign people in and get moving… you couldn’t get anything done.” In fact, keeping programs under 20 clients seemed to be perceived as the ideal among the majority of our participants. Others stated, “Fifteen is the max that would be effective,” and “No more than 12 people in a group, the smaller the group the better.”

Program Duration

Similar to group size, when asked about ideal length of time for intervention, there was no universally agreed upon specific number of weeks for program duration. However, participants all expressed the belief that a greater number of sessions and longer treatment duration were better, given what BIPs were trying to accomplish:

We have to break through those barriers, get them to be willing to change – it is a lot. I think programs should be longer. I couldn’t give you a good estimated guess but I’m going to say, at least, 8 months, 9 months.

Longer programs were, therefore, seen as having more potential to promote behavioral change:

It takes almost 8 weeks for an offender to finally say "Ok, I hit her." The longer you are with the offender, the more change can be possible. So if it takes 8 weeks to admit I did something then you only have another 8 weeks, you can’t change somebody’s behavior in that short of a time period, especially if this behavior has been going on for years.

As another stated, “I know people argue with the fact that 24 weeks is long. I don’t think it is long enough really.”

Some participants expressed a strong opinion that there should be a definitive minimum number of weeks to create the opportunity to elicit some effect. For example, one participant stated, “Program duration must be a minimum of 10 weeks.” Another similarly expressed, “I think it needs to be more than 16 classes.” The majority, however, were not able to provide an exact length of time that ideally would be the most impactful. As another participant stated, “The California system automatically mandates into 52 weeks of classes. Would longer periods of time help our clients? [For] Some yes, some no.”

Facilitation

Participants described two key “good practices” related to BIP facilitation. First, participants felt BIP groups needed to be led by two individuals. Second, participants also believed that individuals facilitating BIP groups needed to have some qualifications for working with domestic violence.

Co-Facilitation

Participants expressed the belief that BIPs should be co-facilitated, or rather led by two individuals as opposed to one, “I think two is better because you can see different things. You are actually literally seeing a different perspective. You can debrief, so you are not just getting your own perspective.” Having two people co-leading groups, it was believed, allowed facilitators to better manage the group and attend to clients. It also allowed those running the groups to gain greater insights into their clients:

You need two facilitators; you need someone watching the group. You can’t always do that if you are focused, but if you have two people, one can watch and observe what is going on and that is oftentimes when you get the real kind of confession of what it is because you’ve caught them reacting.

Some participants felt that ideally such teams should be co-ed, with both a male and female facilitator, “I think an ideal program has a balanced team. I think you have a balanced male/female team.” Those individuals who preferred co-ed teams expressed the belief that such teams were important for ensuring that BIPs avoided the pitfall of subtly reinforcing the very behaviors they were attempting to change, “I always felt that was really important, the male/female co-facilitators to do the group. So it doesn’t turn into a boy’s club.” They also viewed co-ed teams as a more effective mechanism for modeling desired behaviors, “Having a male and a female is ideal because you can role play, you can be a couple working together that shows them, that models the behavior.” Furthermore, these participants believed co-ed teams allowed clients to experience positive, respectful interactions with women:

Having a woman gives them the opportunity to have interactions and exchanges with a female that are positive because she wants to help him improve, and maybe that’s something that he doesn’t have, or hasn’t had a lot in his life, so to have a positive interaction with a female.

Training

Participants agreed that BIPs should be led by individuals who at minimum have some qualifications for working with domestic violence, “All facilitators should have domestic violence training… every facilitator should [be] required to learn about domestic violence.” As another participant stated, “How can you be a facilitator for men who batter if you don’t know anything about domestic violence that you don’t understand the whole power and control?” However, participants differed on what kind of training they believed qualified facilitators to work with domestic violence. Most equivocated, stating that it did not matter what formal education an individual had, so long as they were well equipped to do the work, “If program [facilitators] are well trained… then I think we have the best chance possible to see the best results possible, whatever that might look like.” Though a few believed that some form of graduate training in social work or therapy was preferable, “I guess ideally you want to have someone who’s licensed, either a Masters in social work or at least a licensed therapist of some sort, a psychiatrist. Somebody who’s had some type of higher education in therapy.” Another described how she felt that facilitators needed to have gone through specialized education in treating batterers, “Well I think that we should have people who are qualified teaching the class. I think that’s very important. Qualified, as somebody that has specialized in batterer intervention.”

Program Approaches for Promoting Change

Participants described three intervention approaches that at minimum they thought were needed for promoting change; programs should (1) confront or challenge their clients on their behavior, (2) promote an environment of safety and openness, and (3) strive to adapt to clients.

Challenge Behavior

Participants agreed that a good BIP was one that confronted clients, “The men need to be challenged; the thinking needs to be challenged.” They described various strategies that used in group and designed to challenge clients on their behavior. One way was to consistently get clients to focus on themselves, as opposed to their partner, “I’m not going to argue with him about her. We consistently bring the focus back on to him, his feelings, his thoughts, and his reactions.” This also helped to challenge victim blaming by redirecting clients and addressing their behaviors, “They come in here and I say ‘I don’t want to hear about her. I want to hear your reaction to her. I want to hear how you think. And I want to challenge how you behave towards her.’” Similarly, not permitting the use of derogatory language when clients spoke about their partner was another strategy, “For the individual who is coming to that group… you are not going to call her names, ‘Well that bitch this and that…’ No, we don’t talk like that.” This strategy helped call attention to, and challenge, clients deeply embedded gendered thinking, “One of the things in a group is you don’t call your partner ‘her,’ not ‘she,’ not ‘my woman.’ Use her name. Once you don’t use her name you “thing-a-fy” her. It’s to call people out on it.” Another strategy was to reinforce to clients that their behavior was a choice:

They will say “She pushed my buttons, I lost control. Once that happened I couldn’t pull back, I just went there.” And I say, “If you can’t stop your violence, then you need to be institutionalized. If you won’t stop it, then you need to be in jail.” So it is either if you can’t or you won’t, you could stop it if you chose too.

This approach, likewise, challenged victim-blaming, but also made it clear to clients that they alone were responsible for the actions that landed them in group.

Participants emphasized that BIP sessions needed to encourage clients to actively challenge each other as a part of group process. In fact, there seemed to be implicit agreement among participants that group work was preferable, as opposed to individual counseling or therapy. As one participant described:

My experience as a private therapist is I feel the group process is much better for batterers than one to one therapy. In a group, the clients [begin] to understand that someone knows… how they behave. The guys that have been there longer will say to newbies coming in ‘Oh yeah, I felt that way too, just wait. Keep quiet and listen.’ They have begun to really look at themselves.

Group work therefore maximized the opportunities to confront clients on their behavior, “It’s easy for an individual working with a man to collude with him on the small things... The group is vital because it gives the man an experience of [having] somebody besides the therapist to challenge them.” One participant described how she had the group provide client feedback and suggestions rather than giving it directly herself:

When a client says something ridiculous… you put it in the hands of the group. How would you have responded to that? Just by hearing stuff from other men that have shared the same problem, it is a way to get through to them on certain things.

Peer confrontation through group work was therefore viewed as a way to promote accountability among clients, “It is really important that this be group work, because the men learn to hold each other accountable.” Peer confrontation also helped reach those clients that showed resistance to the facilitator, “In group, it’s easier for them to hear it from another guy than it is from the facilitator. You know, because I’m getting paid to do it.” And thus, it helped to confront clients who were not yet taking responsibility for their behavior and who might be having a difficult time admitting what happened, “The guys are really good at confronting [those] that are in the denial part of it. They really learn from one another. It is outstanding for accountability.”

Promote Safety and Openness

Participants believed that in addition to challenging clients, a good BIP balanced accountability with an environment of safety amongst their clients, “A good program challenges the men that they have but also understands that it comes back to [safety].” Fostering a “safe space” where clients were encouraged to open up, therefore, was one of the goals for BIPs, “One of the biggest goals is to make stuff ‘talk about-able’ – because if they don’t talk about in here, they can’t change it.” One way this was achieved was through having facilitators who were able to address clients in a non-judgmental and understanding way, “It comes down to facilitation; clients need to feel not judged. ‘Yes I understand why you would do this, but let’s look at other ways of handling this, because these are the consequences if you continue down this path.’” Modeling empathy and respect were seen as key strategies to engage clients and encourage them to open up and share:

When clients don’t feel respected, that comes across. You have to show and demonstrate respect. I hate the behavior, but I respect you as a person. And I’m not going to be disrespectful. It is going to be a safe place. You can talk about this stuff. They will not open up at all if they feel you are judging them.

Another way BIPs promote safety and openness is through allowing clients to take leadership and provide, when appropriate, some responsibility for the direction for the group, “I think an ideal group definitely involves process of the men for the men, letting the men talk to each other without a facilitator.” Giving clients the opportunity to interact with each other in a way that minimized facilitator involvement helped to build trust among the group; it allowed members to see that they could share without fear of judgment, “The number of times men have said to me, “Here I’ve just said something that I’ve never said out loud to anyone else and nobody spit on me, went ‘Pfft!’ and walked out of the room!” It also promoted honesty by showing clients they were not alone. One participant described what the client’s rationale might be that lead to openness in group:

‘I haven’t talked about it and don’t want to, [but] the guy across from me just talked about it – like oh my god! I thought that, I felt that, I’ve done that and he’s talking about it. That automatically gives me permission to talk about it.’

Similarly another participant stated, “A lot of people think ‘I’m alone in this.’ These people are not openly saying to friends and family, ‘I beat her up, can you help me?’ So it gives them an opportunity to share that.” Allowing clients to interact on their own terms therefore minimized resistance and encouraging honest group participation.

Adaptability

Participants also expressed the belief that BIPs should be able to adapt to clients, both in terms of addressing clients’ needs and tailoring programs to diverse client populations. Most agreed that it was important for BIPs to have a curriculum to work with; however, participants felt strongly that imposing a strict curriculum on clients was largely ineffective, “It’s more productive for us to talk about what they need to talk about on that particular night. I’m pretty sure we cover [the curriculum] by the time they leave... We just don’t do them [the lessons] in order. Thus groups needed to be flexible enough to adapt “in the moment” and address whatever immediate needs clients might have, “You have to really hone in and look at your group and identify what each of them really needs that night, so allowing the guys to come in and then kind of letting it flow.” Participants reported using strategies such as “check-ins” at the beginning of class to be able to assess how clients were doing each week and help frame the structure of the group that evening:

I don’t come in with a preconceived idea on what is going to happen in group tonight. I have no idea until check-in [process during which clients each describe their current situation and concerns at the beginning of the group session] is done. I always have something I can do that is structured, but if they bring enough, I don’t need it.

Such strategies provided cues to facilitators as to what clients needed to discuss at any particular time, “Whatever their check-in is that day that is where the class goes. If you have a check-in and nobody had issues – that is when you can use different exercises that are in our curriculum.”

Additionally, participants believed that BIPs need to be able to adapt to and meet the demands of different clients. For example, being able to address the needs of different offender types, such as repeat offenders versus first time, and self-referrals versus court mandated individuals, was viewed as important. As one participant stated, “If there is a repeat referral after the participant has completed the initial program the participant must attend a more intensive and long-term program.” Another participant stated, “If we had a group that we could put the men with the different circumstances in, if we had a way to meet the needs of [self-referrals] it would be awesome.” Additionally, BIPs need to consider the learning capacity of clients in their work, “I think some offenders will learn in a different style than other… But I think a lot of times the groups are the same. It should not be cookie cutter across the board because we all learn differently.” Furthermore, participants saw the need to adapt programs so that they were culturally appropriate or, rather, addressed abusive behaviors within the context of different cultural systems:

If your culture shapes your attitudes about women, relationships, and how a man acts in a relationship, then those things have to be directly addressed in order for those behaviors to change, and then reframed within the context of the culture, so that the new behavior becomes more attractive than the old pattern.

Discussion

The results of this analysis found that our participants described key components necessary for a good BIP in ways that reflected their experiences and interactions with IPV and BIPs. Participants described the structural components of groups, and discussed what they believed to be the most appropriate duration of time for program engagement, as well as a group size that would optimize the potential for program impact. Participants also described characteristics of BIP facilitation, including what kinds of training they believed facilitators should have and their preference for what BIP facilitation teams should look like. Furthermore, participants described what they saw as the need to balance an environment of accountability and safety for clients, while also adapting to meet varying needs and cultures.

While participants equivocated on the exact size and length that would be ideal for BIPs, they nonetheless articulated the belief that smaller group size and a longer number of weeks of group sessions increased the potential for effect on clients. Few studies have assessed the relationship between program duration and recidivism or re-assault, with mixed results (Gondolf, 1999; Knowles, 1984; Maiuro et al., 2001; Rosenbaum et al., 2001; Rosenberg, 2003; Stuart, 2005). Furthermore, the number of sessions for BIPs is often arbitrary, and differs greatly from state to state based on standards (Price & Rosenbaum, 2009; Rosenbaum et al., 2001). Our findings indicate that professionals working in or with BIPs support the notion that there may be a greater opportunity for sustained behavioral change in longer, more intensive programs. However, a greater examination of the perspectives of other IPV professionals working on the ground with BIPs, BIP clients, and client partners is needed to provide a better sense of whether or not longer program duration /greater number of sessions ultimately impacts more intermediary outcomes, such as clients’ beliefs and attitudes.

We also found that a majority of participants felt that BIP facilitation should be undertaken by teams of two facilitators who have some training or qualifications for working with intimate partner violence. Additionally, for some, co-facilitation was ideally comprised of male/female teams. Given the limited information in the literature on facilitator gender, it is unclear whether or not facilitator gender has an impact on client experiences (Bailey et al., 2012; Roy et al., 2013; Tyagi, 2006). What is perhaps more important is our participants’ views on the qualifications necessary to facilitate a BIP; and in specific, that most did not believe that a formal, specialized education in battering was necessary to do the work.

Reviews of BIP standards have shown that qualifications for BIP facilitation vary from state to state (Austin et al., 1999; Maiuro & Eberle, 2008; Price & Rosenbaum, 2009). At a minimum, most states require BIP facilitators to be violence- and alcohol- and drug-free (Maiuro & Eberle, 2008). However, despite a trend toward professional licensure, many states (40 %) still do not require even a bachelor’s degree to be credentialed as a BIP facilitator (Maiuro & Eberle, 2008). Furthermore, at least 13 % of programs allow ex-batterers to act as facilitators with the only known requirement being that they must be violence-free for at least a year prior (Price amp; Rosenbaum, 2009). Our study begs the question then of what constitutes qualifications for BIP facilitation, and furthermore, suggests perhaps that the qualifications for BIP facilitation do not necessarily need to include formalized education. This is not to say that facilitators should not be keenly sensitive to the complexities of domestic violence, but rather, given this finding along with the fact that our participants saw empathy and respect as key facilitator strategies, a consideration might need to be given to what personality traits, or characteristics are desirable for BIP facilitators.

Our participants also emphasized the importance of facilitators using various strategies which rely on the group dynamic to promote personal accountability, demonstrate respect and safety to encourage honesty, and enhance the relevance and credibility of the group for clients. First, our participants described relying on clients and facilitators in group to challenge each other on their behavior. Such an approach is characteristic of the Duluth model (2011), and thus, it is perhaps not surprising that our participants described using this strategy as our programs both used some variant of this model. Similarly, other research has found confrontation to be an essential part of the BIP process (Chovanec, 2009, 2012; Parra-Cardona et al., 2013; Scott & Wolfe, 2000; Silvergleid & Mankowski, 2006). Others have shown that clients themselves have described being held accountable by other group members and facilitators as one of the more helpful components of BIPs, (Chovanec, 2009, 2012; Parra-Cardona et al., 2013; Scott & Wolfe, 2000; Silvergleid & Mankowski, 2006) and that such client interactions possibly deflected anger and defensiveness, and engaged clients in a shared learning process. Critics of the Duluth model (2011), however, have argued that a confrontational approach may inadvertently decrease the therapeutic impact of BIPs by increasing resistance among clients (Dutton et al., 2007; Mankowski et al., 2002; Milner, 2004; Murphy & Baxter, 1997; Musser et al., 2008). Thus, while our study supports findings that suggest that challenging clients may be useful for engaging BIP group members in the work of reframing beliefs, attitudes, and ideas about their relationships and experiences, it is unclear to what extent this approach may also conversely inhibit clients and prevent them from engaging in the group process.

This point is particularly salient given that participants also described the need to use the group to model and demonstrate respect and safety, and promote openness and honesty among clients. The issue of safety, as related to promoting openness among clients in BIPs, is well documented in the literature. Campbell et al. (2010) noted that BIP clients are often reluctant to disclose violent behaviors due to fear of shame, embarrassment and judgement; furthermore, they found that trust and confidentiality played a critical role in help-seeking among batterers. Similarly, Chovanec (2009) found that BIP facilitators felt that validating clients’ experiences, and being sensitive to shame or blame, helped engage clients in the group process. Still others have recognized the importance of promoting non-judgment in groups as key to encouraging clients’ openness (Bailey et al., 2012; Campbell et al., 2010; Chovanec, 2012; Parra-Cardona et al., 2013; Rosenberg, 2003; Scott & Wolfe, 2000; Silvergleid & Mankowski, 2006). Our work supports this literature and furthermore suggests that modeling safety and respect may help to minimize client resistance by reducing clients’ fear of judgment, and may increase clients’ willingness to engage in the process of behavioral change BIPs are trying to promote. However, this study also begs the question of whether or not confrontation can be balanced against the promotion of safety and honesty, and if so, how? More research is need to understand the role that confrontation plays in the BIP group process, and whether or not there are effective ways to challenge BIP clients in ways that will help to reduce perceived power differentials between facilitator and client, and promote engagement among clients through the utilization of confrontation as a learning exercise rather than a rebuke.

Third, our participants also believed that allowing the group to share their narratives and experiences promoted a sense of applicability to, and relevance for, the clients’ personal lives and circumstances. Allowing clients to create their own space in group through “client initiated” or “client driven” group discussion has also been recognized as an important component for BIPs (Campbell et al., 2010; Davis et al., 1999; Parra-Cardona et al., 2013; Rosenberg, 2003; Scott & Wolfe., 2000; Silvergleid & Mankowski, 2006). It should be noted here that affording clients some freedom in group does not equate to a “free for all.” Instead, the belief is that clients learn through participating in a structured context with guidance from a knowledgeable facilitator. This process correlates with models of adult learning and progressive education (e.g. Knowles, 1984) that argues that learners, particularly adult learners, tend to be more receptive to learning that is active rather than passive, and that actively working through skills and information, rather than merely receiving teaching, allows adult learners to find salience with their own lives and experiences.

Each of the three strategies our participants endorsed suggests that perhaps, instead of focusing on what models are most appropriate for BIPs, more attention needs to be paid as to how the group process works to support client behavioral change. However, the literature testing the efficacy BIP models has been largely inconclusive (Alexander et al., 2010; Brannen et al., 1996; Coulter et al., 2009; Dunford, 2000; Dutton et al., 1997; Eckhardt et al., 2008; Feder et al., 2002; Gondolf, 1999, 2000, 2004; Gordon & Moriarty, 2003; Jones et al., 2004; MacLeod et al., 2009; Morrel et al., 2003; O’Leary et al., 1999; Pascual-Leone et al., 2011; Saunders, 1996; Stuart, 2005; Taylor et al., 2001). Other smaller studies which have examined the perspectives of clients’ have found that often what clients view as helpful about their BIP experience is the group process or dynamic, rather than the specifics of what is taught in group (Chovanec, 2012; Parra-Cardona et al., 2013; Rosenberg, 2003; Silvergleid & Mankowski, 2006). This is not to say that programs should not address issues related to gender ideology or power and control, or that program content does not matter; rather, given that studies also show that in general BIPs exhibit low attrition rates (Bennett et al., 2006; Brown et al., 1997; Buttell, Frederick et al., 2002; Buttell & Carney, 2008; Cadsky et al., 1996; Chang et al., 2002; Daly et al., 2000, 2001; DeHart et al., 1999; DeMaris, 1989; Grusznski & Carrillo, 1988; Hamberger et al., 2000; Jewell & Wormith, 2010; Rooney & Hanson, 2001; Yarbrough & Blanton 2000), what perhaps is needed is a better understanding of how BIPs engage clients in the group process. More observational studies that document and describe how the strategies identified in this work operate in “real time” are needed therefore in order to understand how the BIP process can be best facilitated to promote client growth, and ultimately change.

Finally, we found that participants believed that BIPs should be tailored to meet the needs of diverse clients. The need for BIPs to expand beyond a “one size fits all” approach is not a new concept, and others have criticized current models due to the lack of relevance they may have for minorities and other diverse populations (Becker et al., 2012; Maiuro & Eberle, 2008; O’Leary et al., 2008; Saunders, 2008). A growing body of research, therefore, has sought to support the notion that BIP groups should be culturally competent or tailored specifically to different cultural groups (Buchbinder et al., 2008; Buttell, 2005; Gondolf, 1988; Parra-Cardona et al., 2013; Saunders, 2008). However, less information exists on what alternative approaches for different batterer types (Hamberger et al., 2009;Saunders, 1996; Whitaker & Niolon, 2009), and virtually no research has examined how BIPs meet the different learning needs of clients. Thus, in addition to supporting the need to understand how BIPs can best be tailored to be cultural competent, our study also suggests that observational research is needed that specifically examines how clients respond to and engage with different BIP models, including both psycho-educational and cognitive behavioral therapy programs. Additional research is also needed to determine if certain models work better for different types of clients than others, as well as how facilitator characteristics and processes related to group dynamics impact client outcomes.

Limitations

This study has several limitations that deserve mention. First, the study is focused specifically on BIPs for male perpetrators of IPV against female partners, using models that are fairly similar in structure. Thus, our findings may not be relevant or generalizable to other kinds of BIPs that address different populations (e.g. groups for LGBTQ individuals, female perpetrators, etc.), or BIPs that utilize other models for intervention. This study was also conducted in a single county. Our findings may then be isolated to this region and the various legal, social, and political influences specific to our region. However, we have reviewed our findings with BIP experts from other states and counties. While we heard some differences of opinions regarding gender and number of facilitators, these reviewers corroborated our findings regarding smaller groups and longer sessions, as well as the importance of using group approaches to enhance learning objectives. Our sampling strategy and number of key informants also does not allow us to make comparisons between categories of participants. There may be differences in perspectives between the community advocates and counselors and professionals in the law enforcement and legal fields that we were not able to capture. We also chose to stop subject recruitment at 36 key informants. The addition of more informants—particularly those with positions or roles that had less representation in our sample—could have elicited new or different perspectives. However, we noted thematic saturation as early as the 10th interview and, thus, felt comfortable that the perspectives we had obtained represented the most robust themes.

Conclusion

Despite these limitations, our findings have policy, service and research implications. These findings highlight the need to consider BIP structure, facilitation and approach in addition to content, curriculum or philosophy before any assessments of BIP impact and/or quality are derived. Additionally, these findings offer some insight to participants’ perceptions regarding how change may occur among BIP clients—that these changes may be gradual and take time, may benefit from the insights and skills of more than one facilitator, and are more likely if inspired by client’s internal motivation rather than external inducements. Each of the identified components of a good BIP in our study can be a focus for future examination to assess whether and how these elements contribute to change among BIP clients. Finally, these findings may inform policy makers and community advocates as more states and regions develop standards and requirements for BIPs.