Voluntary, Survivor-Centered Advocacy in Domestic Violence Agencies

Voluntary, survivor-centered advocacy is a model of practice used in domestic violence organizations; however, more information is needed from the perspective of survivors on how to best facilitate survivor-centered approaches in a voluntary service format. This qualitative study used a thematic analysis to uncover core advocacy approaches from 25 female-identified survivors dwelling in domestic violence emergency shelter and transitional housing programs in two states. Themes revealed that three core approaches aid a voluntary, survivor-centered advocacy model: 1) Establishing a safe base for support, 2) Facilitating access and connection, and 3) Collaboration. Advocacy approaches that emphasize safety, mutuality, and availability of support best engage survivors in voluntary services to address needs and meet goals. Use of a strengths-based approach, psychoeducation, and resource-building contributes to the social and emotional well-being of survivors. Findings indicate community DV advocates should use adaptable advocacy models aimed at service access, connection, and collaborative resource acquisition. Voluntary, survivor-centered models use principals of trauma-informed care, though more widespread use of trauma-informed care (TIC) in voluntary services are needed. Advocates need organizational support to meet survivor needs. Implications for research include the need for fidelity studies and longitudinal research.

Services for domestic violence (DV) survivors have a long history of using empowerment and agency-building strategies that help survivors to advocate for needs and rights (Cattaneo & Goodman, 2015;Wood, 2015). Advocacy is a primary direct service model used with survivors in traditional DV services. At their core, advocacy services represent a departure from traditional case management in social services to an empowerment-based approach to work with DV survivors on their defined goals rather than providing a prescriptive, one-size-fits-all approach ( Goodman et al., 2016b;Sullivan, 2018;Wood, 2014). Per federal policy (The Family Violence and Prevention Services Act of 1984 [FVPSA]) and in alignment with an empowerment perspective, advocacy services are typically offered on a voluntary basis, without service use as a condition for shelter or other resource use (Nnawulezi et al., 2018). A growing body of research has recognized advocacy for DV and sexual assault survivors as a distinct service model ( Davies & Lyon, 2014;Goodman & Epstein, 2008;Wood, 2014); however, more information is needed to enhance voluntary, survivor-centered advocacy practices (Davies & Lyon, 2014;Goodman et al., 2016c;Nnawulezi et al., 2018). Given the centrality of survivor voices in the work of DV advocates, more information is needed on the perspective of DV survivors about what they need and want from their work with advocates, in order to further develop best practices. To amplify survivor voices in this area and enhance the efficacy of voluntary, survivor-centered models in DV agencies, the current study used thematic analysis to understand what advocacy approaches aid a voluntary, survivor-centered model from the perspective of those engaging in services. Twenty-five DV survivors who were actively participating in voluntary advocacy services at emergency shelter or transitional housing programs participated in semi-structured qualitative interviews to share their perspectives about enhancing services.

Literature Review
DV is a pervasive public health problem. The Centers for Disease Control and Prevention (CDC) National Intimate Partner and Sexual Violence Survey (NISVS) found that 37.3% of women reported experiencing violence by an intimate partner during their lifetime (Smith et al., 2017). DV victimization can result in physical injury and chronic physical health problems, such as headaches, asthma, and diabetes, and may lead to hospitalization or death (Breiding et al., 2014). Mental health impacts from DV include post-traumatic stress disorder (PTSD), depression, anxiety, insomnia, substance abuse, and trauma symptoms (Campbell, 2002). DV can also negatively impact economic stability, housing stability, and outcomes for children, and is a leading cause of homelessness in the United States (Jasinski et al., 2002;Klein et al., 2019).
DV shelters, non-residential services, and housing programs are the mainstream refuge for survivors facing safety and economic challenges (Baker et al., 2009;Panchanadeswaran & McCloskey, 2007). The National Network to End Domestic Violence (NNEDV, 2017) estimates 1,910 DV-focused agencies exist in the United States to meet the needs of survivors and their children. An estimated 72,959 survivors are served in a single day, with 41,195 through shelter or transitional housing and 31,764 through non-residential care (NNEDV, 2017).

DV Advocacy Services
DV programs aim to increase survivor social and emotional well-being by fostering self-efficacy and increasing access to resources and support, often through advocacy services (Sullivan, 2018). Advocate is the common term for a direct service worker in the field doing work that is generally not clinical in nature. Advocates work with DV survivors to help increase safety, meet social and economic goals, access resources, and promote healing from trauma (Goodman et al., 2016b;Rivas & Vigurs, 2018;Sullivan, 2018). Advocates frequently come from social work or other helping professional backgrounds Wood et al., 2019). Advocacy contributes to increases in outcomes related to survivors' well-being (Sullivan, 2016). Well-being includes emotional, social, and physical components such as perceived quality of life, feelings of support from others, and physical well-being (Sullivan, 2016). Previous research indicates that by providing information; offering encouragement, supportive counseling, and safety planning; and increasing access to resources and social support in advocacy services, survivors reported increases in safety, resource acquisition, and emotional and social well-being (Constantino et al., 2005;Haj-Yahia & Cohen, 2009;Lyon et al., 2008;Sullivan & Bybee, 1999;Tiwari et al., 2010). Additionally, advocacy has been shown to decrease negative outcomes such as subsequent experiences of violence and negative mental health symptoms (Bybee & Sullivan, 2002;Sullivan & Bybee, 1999). Findings from Wathen and MacMillian (2003), Sullivan and Bybee (1999) and Ramsey et al. (2016) indicated that advocacy significantly reduced subsequent victimization in DV survivors up to two years after exit from a shelter. Goodman et al. (2016b) found that the alliance between the advocate and survivor was essential to increasing well-being. In a sample of 370 survivors, a stronger alliance predicted decreased depression and PTSD through empowerment-based safety.
While shelters and housing programs remain the central hub of DV services, advocates also work in other institutional settings, such as non-residential services, courts, medical offices, law enforcement, and counseling agencies (Allen et al., 2004). While variations in advocacy models exist, the 2010 update to the Family Violence Prevention and Services Act (FVPSA) clarified that DV shelter supportive services receiving federal funding, like advocacy or counseling, must be available to service users on a voluntary basis, and that participation in services must not be used to determine shelter eligibility (U.S. Department of Health and Human Services, 2012). A voluntary service model, as outlined by FVPSA, provides respect for survivor autonomy and allows room for individualization based on each person's unique circumstances and needs (Missouri Coalition Against Domestic & Sexual Violence, 2012). Voluntary and low barrier, or more accessible, services are inherently more survivor-centered than mandated service models, and are linked to increased survivor autonomy and empowerment (Nnawulezi et al., 2018). To enhance the effective implementation of survivor-centered voluntary advocacy services, an understanding of the perceptions of survivors using services is critical.
Survivor-Centered Advocacy. Increasingly, survivor-centered advocacy, a service model based on a feminist, empowerment approach to promote individualized survivor well-being, is endorsed for voluntary DV services (Davies & Lyon, 2014). Survivorcentered advocacy acknowledges the variety of needs of individual survivors and is sensitive to the diverse needs of survivors and their families. This approach allows survivors and advocates to shape services based on their own goals, with an emphasis on partnership between the survivor and advocate (Goodman et al., 2016c). While few established models of advocacy have been empirically tested, available research suggests individualized advocacy services are effective in addressing social and emotional support needs (Allen et al., 2013;Bybee & Sullivan, 2002;Goodman et al., 2016b). Two examples of survivor-centered advocacy are described below.
The Community Advocacy Project (CAP) was designed to be a survivor-centered, short-term, intensive service model (Allen et al., 2004;Allen et al., 2013;Bybee & Sullivan, 2002;Sullivan & Bybee, 1999) and helped to establish evidence for core advocacy services. CAP was found to be effective in increased outcomes such as safety, social support, quality of life, and well-being, especially when coupled with increased time with an advocate, and the ability to meet survivors in a safe convenient location (Allen et al., 2004;Bybee & Sullivan, 2001). A process evaluation of CAP from Allen et al. (2013) highlighted that critical program components included orientation to the whole person, unconditional validation, and information provision and action (Allen et al., 2013). The CAP model has been adapted cross-culturally and has been disseminated in areas of Latin America. Findings in Mexico have suggested similar efficacy after cultural adaptation .
Another model of survivor-centered practice is the Victim Defined Advocacy (VDA) framework developed by Davies and Lyon (2014). This framework centers on survivor autonomy and empowerment, and the assessment of both batterer-generated risks and lifegenerated risks. These risks are posited to work across levels and systems to inhibit survivor's quality of life and safety (Davies & Lyon, 2014). Other key components of survivor-centered practice include advocates working with survivors to build partnerships, review risks, and identify options based on relevant priorities, decisions, and dangers (Davies & Lyon, 2014;Goodman et al., 2016c).
Theoretical orientation. A feminist empowerment frame is a central theoretical construct in voluntary DV advocacy, rooted in the idea of undoing the abusive partner's use of power and control. Empowerment approaches seek to give power back to those who have been oppressed and are closely linked to feminist and critical interventions (Dominelli, 2002), as well as crisis intervention models that emphasize personal agency and decision-making (Roberts, 2007). Cattaneo and Goodman (2015) describe empowerment-based practice for DV survivors as an interaction with the social world where a person is able to act on meaningful goals, draw on support to meet those goals, and reflect on the extent to which the action taken results in desired progress. Qualitative interviews with women staying in a shelter suggest that interventions are more meaningful when they seek to empower survivors through mutual power-sharing, information-giving, and advocacy (Haj-Yahia & Cohen, 2009). Voluntary and survivor-centered models of advocacy are closely theoretically linked to an empowerment perspective, but more information from the express perspective of survivors using DV services is needed to guide effective implementation (Goodman et al., 2016c;Sullivan, 2018).
Need for the current study. While many DV agencies strive for a survivor-centered voluntary service model, there is little guidance from the survivor perspective about how to best apply this advocacy approach. Research in implementation science points to key characteristics of health and social service interventions that are likely to be successfully implemented with fidelity and longevity (Damschroder et al., 2009). Evidence finds stronger implementation outcomes for interventions when staff and agency leaders have a positive sense of the source and quality of the evidence surrounding the intervention, when they have a positive sense of the perceptions of clients related to the intervention, and when the intervention includes a clear set of manageable and defined tasks (Damschroder et al., 2009). Research supports the efficacy of advocacy service models such as CAP (Allen et al., 2004) and highlights the importance of the advocate-survivor alliance (Goodman et al., 2016c). However, perceptions of voluntary survivor-centered advocacy services from the survivor perspective are lacking, and less is known related to the defined tasks of effective advocacy services. The survivor perspective of advocacy skills and approaches is crucial to help advance the successful uptake, implementation, and subsequent evaluation of a survivor-centered advocacy model. Thus, to amplify the voices of survivors of violence, the current study used qualitative thematic analysis (Braun & Clarke, 2006) of narrative data from DV survivors at four agencies using a voluntary service model to explore what advocacy skills and approaches help survivors meet their needs.

Method
The data from this study are a component of a larger study that aimed to understand the experiences of services for DV survivors using shelter and transitional housing programs (Wood et al., 2017). The current study focused on survivors' perspectives of voluntary, survivor-centered advocacy approaches. The research question guiding the analysis was: "What approaches aid a voluntary, survivor-centered advocacy model?" Twenty-five women participated in this study. The lead author recruited research participants from four non-profit organizations focused on serving survivors of domestic violence across two diverse states (one in a Midwestern state and three in a large southwestern state). The participating organizations provide counseling and supportive services, shelter and other housing programs, and advocacy services for survivors. Organizations selected were domestic violence focused-agencies known to the first author. All organizations identified as using a voluntary service model. Participants included residents and former residents of shelter and transitional housing programs, all of whom had participated in advocacy services. The lead author worked collaboratively with agency staff in shelter and transitional housing programs to recruit participants using flyers and to schedule interviews with interested current and former residents who identified as DV survivors. An effort was made to recruit both shelter and transitional housing participants to gather a range of duration and intensity in the advocacy relationship. Recruitment was open to participants of any gender and yielded a sample of all female-identified. Table 1 describes selected characteristics of research participants, including the type of services received by participants, age, racial/ethnic background, and education.

Data Collection
In-depth, in-person interviews were the primary data collection strategy. All participants received written information describing the study and information important for informed consent to participate, in addition to cash compensation ($20) for their time and expertise. Interviews lasted on average an hour and were digitally recorded and transcribed. No identifying information was collected. The lead author, a social worker with extensive experience working with DV survivors, interviewed all participants inperson and on-site in a private area.
The semi-structured interview guides designed for this study aimed to elicit participant experiences through open-ended questions about survivors' experiences with advocates, their needs and preferences for advocacy services, and their observations and opinions of advocacy approaches. The interview protocol included broad open-ended questions about the experience in shelter, including: "What is it like to live here?"; "What do survivors need when they come to (agency)?"; "What was it like to work with your advocate?"; "How did you and your advocate approach your work together?"; "What support did you need from your advocate?"; "What would you change about services here?" This study was approved by and complied with the first author's Institutional Review Board at the time, Texas State University. Researchers assigned pseudonyms to each participant for the purposes of analysis and dissemination of findings. While the interview protocol did not ask about past abuse, interviews nonetheless held the potential of eliciting painful or traumatic memories of abuse and violence. At any point during the interview, if the participant experienced discomfort, they were free to pause or end the interview, though this did not occur in any interview. Given the method of recruitment and collaboration with program staff, all participants who expressed immediate needs were referred to an on-call staff member who could address any concerns resulting from the interview. The interviewer distributed a resource list of additional services as needed.
Data Analysis. Researchers used thematic analysis approaches (Braun & Clarke, 2006;Clarke & Braun, 2013). The first and third author first read through the data for initial familiarization and beginning identification of themes. Then they independently organized and developed initial codes, coded data, beginning with open, line-by-line process coding of full interview transcriptions to search for and categorize themes related to engagement with advocacy services (Charmaz, 2006;Clarke & Braun, 2013). Next, the researchers reviewed and mapped themes to examine relationships among the themes. A total of seven themes, with 18 subthemes, were in the final codebook. A final phase of analysis involved the reassembling and rich description of major themes identified during the initial phases of coding (Braun & Clarke, 2006;Saldaña, 2013). The researchers engaged in memowriting and peer debriefing between coders throughout the analysis process. Regular peer debriefing was used to resolve coding disputes and share analytical memos. The themes saturated across all participants are presented in this analysis.
The following strategies enhanced the quality and rigor of this study: worthy topic, credibility, audit trail, and peer debriefing (Creswell & Miller, 2000;Lietz et al., 2006). First, survivors' perspectives on and experiences with the advocacy approach in the context of shelter and transitional housing represents a topic worthy of exploration and is relevant to both survivors and social service practitioners. Second, researchers sought credibility through digital recording and transcription, multiple coders, and triangulation meetings. Throughout the process, researchers maintained an audit trail of memos, methodological decision-making and processes, and analysis decisions (Rodgers & Cowles, 1993). Initial findings were reviewed with practitioner collaborators as a form of peer-debrief.

Findings
To better facilitate the use of survivor-centered advocacy, this research explored survivors' relationships with their advocates and how advocates helped to meet their needs and achieve goals using a voluntary service model. These findings underscore the help and potential healing benefit from advocacy services, and conversely, the hindrance of negative interactions between survivor and advocate. The relationship between the advocate and survivor during services is one that at its best, is built from collaboration, empowerment, mutual guidance, and the advocate's skill, warmth, availability, and equitable treatment. The survivors' motivation to work with the advocate to achieve goals also contributes. Conversely, when the advocate and survivor relationship is less than positive, conflictual, or incomplete, there is a lack of fit between needs, approach, and skills. At its worst, the survivor may feel ignored, discriminated against, or actively silenced. Participants in this study indicated that the best approaches for survivors and advocates for a voluntary, survivor-centered model are characterized by: A) A safe base for support, B) Access and connection, and C) Collaboration.
A Safe Base for Support. Survivors emphasized the need they have for advocates to establish an emotionally and physically safe environment. Establishing safety begins with the first interactions with advocates to take care of their immediate physical needs. Commenting on her intake process, Rhonda stated: They just did intake and gave me little bags with snacks in them and gave me my sheets and just calmed me down, told me everything was gonna be okay and just left me alone. Let me be comfortable.
Safety is also clearly signaled by the built environment (e.g., gates and locks on the doors) as well as the structure of human resources and housing staff (e.g., staff on-call around the clock). However, advocates also play an important role in establishing safety and do so by developing kind, caring rapport with clients. Many participants articulated that being able to talk to their advocate and receive empathic reactions were key to feeling not only physical safety but also emotionally safe in their relationship with their advocate. Ruby described how her advocate helped establish safety through rapport-building: Participants also identified the importance of confidentiality and privacy in ensuring their sense of safety. Privacy included the need to maintain control over the speed and pacing of when and how much to share with an advocate. Gracie, for example, remarked, A lot of times, I think, like even with me when I first got here, I didn't feel comfortable talking in front of anybody. I didn't want anyone to know my story.
Not feeling pushed, or even obligated, to share certain aspects of their life was important in boundary-setting and building trust. Makayla recalled early interactions with advocates: Listen, like when the person talks, and just invite them to be open and don't be so pushy like asking them the same question, but re-wording it. I get that a lot, like I'm not understanding what you're doing. I don't know-just don't be so pushy.
Safety is also facilitated by the advocates' use of a non-judgmental and understanding approach, illustrated through their words and actions, and thus emerged as an important need expressed by participants. Understanding means the advocate was aware of the challenges, barriers, and experiences of survivors. Participants expressed the need for advocates to both sympathize and empathize with clients without expression of judgment for past actions. Nicole emphasized that being understanding does not always necessarily mean being overly nice or avoidant. She described her advocate, She's quite possibly the most understanding person I've ever met in my life, but at the same time, she just doesn't take bullshit.
One cue that an advocate may be understanding is if they have shared life experiences that might lead to increased empathy or the ability to connect with survivors, as Makayla noted: The [staff] here-we've been through the same thing. Knowing that there was somebody who went through the same thing, they weren't judging me or looking at me like, 'How could you be so stupid? Why'd you stay for so long? Why'd you have five kids and then leave?' Despite the potential benefit of advocates having a shared background with DV, survivors emphasized that non-judgmental understanding involves an acknowledgment of the diversity of possible survivor experiences and advocates' recognition of their own limited knowledge, understanding, or shared experience. When asked what an advocate could do to better help survivors, Ivy noted that refraining from acting as a judge could help to create a better connection: Remember that you are not a judge. Never to think, hard to say, 'I know what you're going through,' because every single story is different. I've heard someone say it and I feel like-I don't say it out loud, I think it to myself-they're still in the beginning of this journey when they say, 'I know exactly what you're going through,' or 'I've been through the exact same thing,' but I'm like, no, you weren't there that night he da-da-da and I wasn't there that night that he to you da-da-da-da, so no, we aren'tan advocate, we all, need to remember that.
While some advocacy approaches carry the unintended consequence of making a survivor feel judged, survivors recognize and articulate the mixed benefits of these approaches. Advocates can promote safety by explaining the voluntary nature of services. When services are not perceived as voluntary, the impact can be detrimental. For example, Tammi described feeling as if she had to meet with her advocate every day.

It's really stressful for me. I'm having a lot of issues with anxiety and I almost feel like I'm in front of the teacher for an evaluation.
Access and Connection. Access and availability of the advocate helps facilitate connection to engage in work together, which is especially important when there is no mandate to use services. Participants repeatedly asserted that the availability of and access to their advocates was important to both establishing rapport and getting their needs met. Consistent, on-request availability provides survivors the opportunity to reach out to advocates as issues arise and facilitates feeling supported. Participants also articulated the need for advocates to be explicit about what to expect as far as availability. Makayla suggested,

Let them know 'if you need somebody to talk to, I'm here to talk to you. You can talk to me about anything, feel comfortable, just knock on my door and we can talk.
The perception of availability and accessibility is furthered by the presence (physical or electronic) of the advocate and by proactive check-ins to see if clients need support. When an advocate is perceived as hands-off, participants view this as advocates being inaccessible, thus limiting rapport and connection. Not only does a hands-off approach inhibit their ability to interact with their advocates, but some participants perceived this as the mechanism through which their needs are left unmet, and thus they receive lower quality care. As Gracie shared: Yeah, I think they should offer us other counselors instead of the counselors that are here because they're so busy-there's so many people here that they can't see us all and anytime if we have a crisis or are going through something, they're booked out a week in advance, so what are we supposed to do? We have to just deal with it.
Participants were aware of the multiple demands on advocates' time and scheduling. They regarded advocate availability with such importance that they may prioritize other survivors' needs over their own. Cheryl described reducing her use of the advocate so that other survivors had more opportunities:

I'll stay and deal with it, but a lot of the women here need more help than what I need and so I stand back and let my advocate who's-my advocate's not just my advocate, she's other people's advocate-so I try to stand back and let her do what she needs to do for them because they obviously need a little bit more attention than I do, and everybody's story's different and situation is different and their ability to work or not work because they're bi-polar or they're just afraid to leave…
Engaged listening, or listening that reflected active understanding of survivor needs, also facilitates access and subsequently, connection. Participants identified a need for their advocates to not only listen to them but to hear them and reflect that in their interactions and engagements. Paula clearly expressed the need for engaged listening to take priority over other tasks: Listen to them. Listen to them carefully. Listen. Carefully. And let them talk. Let them talk. Paperwork-no, don't grab all that first. Let them talk. Listen, let them express themselves. Let them get it out, get everything out. That's a relief coming in places like this. You know what I'm saying? That's a relief coming to a place like this, let me get it out instead of telling me, oh, sign this, these are the rules-no, let me get it out. Let them spill their guts. I mean, for however long it takes, let them talk to you. They wanna tell you and they wanna tell you stuff that they've been wanting to say and then see, they're holding it in and all they do is adding on to the anguish, the depression, the bitterness, until the next place.
Participants indicated that when advocates listen and as Gracie noted "pay very good attention at what they say and don't make them repeat themselves," and reflect that information back, they are able to gather information to better inform their interactions with their clients in the future and promote a sense of alliance. Engaged listening can have both a relieving and an empowering impact on survivors, as Cheryl described:

It just-it opens my heart up and opens my mind, I don't know how to explain that. What does that do? It makes me feel awesome. It makes me so that I can go upstairs, it takes so much away from me to just listen to me that when I go upstairs, I'm not mad at (daughter) anymore, she could be screaming and hollering at me, but I'm not mad at her no more and I can understand where she's coming from a little bit better because at first, we're going at it, we're going at it, and I'm not trying to understand her, I'm trying to get her to understand my point, but then when I walk away and I have a chance to talk to my advocate and she just listens to me and says things like, 'well, how does that make you feel? How do you think that makes her feel?'
In engaged listening, advocates often respond with supportive efforts to help clients generate ideas about potential solutions to meet needs and accomplish goals. Interestingly, survivors also identified the delicate balance between listening and using typical case management skills such as goal-setting and connecting clients to resources. For example, survivors reported differing perspectives on giving and receiving advice from advocates. Some participants wanted advocates to listen and not give any feedback. Others wanted advocates to actively listen so they could provide feedback and input, as Brandi noted:

No, they just need to sit down and listen. Just listen to the clients. Listen to them and hear them real good and then give them some feedback and then they'll still talk. They'll talk some more if you feel like you wanna talk to them, you know what I'm saying? I know you have different appointments-you can't spend the whole day with one client. But at least give them 30 minutes and listen to them and then give them the feedback on what you heard them tell you. And some victims it motivates them, like me-it do. If I'm talking to you and you're my counselor and I'm letting you know I'm depressed today, ok, when I'm depressed and I'm telling you, I want you to be able to give me some input back to let me know, ok, what can I do to make it better.
When advocates are perceived as available and accessible, survivors reported reaching out more frequently to address challenges and needs and engage in services. Joy described how her legal advocate's high level of access helped to build rapport and achieve goals: Collaboration. Safety, access, and connection build the foundation for active, survivor-centered work on needs and goals in a voluntary service model. Advocates not only offer resources and support to survivors, they may also help survivors to create a path forward or guide them to meet goals like exiting sheltering into safe affordable housing, addressing mental health care needs, or removing barriers to resource access. Ivy used an analogy to describe the collaborative work of advocacy as paddling together:

Say you're in a canoe and instead of going like that and shoving somebody out into the middle, give her a nice little kick-off and just let them drift, but if they start going towards the wall, they grab the paddle and let it slide back the right waythat's more them instead of *swoosh*, stop it and turn it around.
Paddling, or collaborating together, represents using flexible skills and approaches to meet goals. One of the primary ways advocates collaborate to help survivors on their journey is through education and normalization. Participants shared how their advocates were able to share information about available resources, common trauma reactions, untangle legal jargon, and breakdown strategies of abusive partners. Providing information and support, especially about the impact of abuse, can help survivors to reduce blame and achieve goals. As Elena noted, psychoeducation normalizes survivors' experiences of the abuse, as well as the process of healing from violence: Whenever I would tell her about the problems I was having with anger and stuff, she was-she would tell me that that was normal, basically and it's nice being able to hear other people having similar situations and then you feel less crazy. I had told her one thing where he was mostly-he would just fuck with my head constantly and that was his game and he did it on purpose and the farther I got away from him, the more I realized examples, would write them down and talk to her and I told her that one time, one example I told her, I just asked him to hit me because I wanted him to get out the anger and she's saying-she made me feel like that was not as crazy as I thought it was at that point because I just wanted him to get it out, basically, and it was very normalizing for me to realize that both I was doing well because she would comment about how I'm doing in dealing with everything and also that so many other people go through these same steps that I was and then when you fall back, she's still there to tell you that you're okay and you can keep on moving forward and also with different kinds of stuff of dealing with the stress.
Advocates also support survivor goals by engaging actively in advocating for clients to help them gain resources and meet goals. Participants stated that their advocate needed to both work for them and with them. In discussing advocates working for them, participants talked about connecting through listening and the philosophy of survivordriven services. For example, Brandi stated, "You work for me, I'm the one that's the victim. Without me, you wouldn't have a job." Specific examples of working for participants include making calls, connecting to other services, lobbying decision-makers at housing and other resources, and motivating participants. Survivors also identified the ability to advocate and vouch for survivors in difficult situations as a component of working for participants. As Cheryl described: It's my hope that my advocate can go to bat for me to Section 8 and tell them, 'look, since she's been here, she has gotten her a car, she's never been late on her rent,' and they can go to bat for me to be able to get into Section 8 instead of me being on housing authority property around all the people that I shouldn't be around because of my drug charge, I'm hoping that she can go to bat for me and get me into Section 8 so that I can be able to pay the lesser rent and take care of the children better.
The ability for an advocate to work for or alongside survivors was perceived as an indication of caring and respect. Furthermore, participants described relief that their advocates were partners and shared the workload of rebuilding their lives after fleeing abuse. Shira shared an example: Well, for instance, for this new place, we've been calling around, me and my women's advocate have been-we got the Section 8 list of apartments and we've been calling around. She would do half of them and I would do the other half just to see if we got anywhere and we came across two places that were accepting people by next month.
Unlike mandated or manualized service approaches, a voluntary, survivor-centered service model takes a flexible approach to advocacy. Alongside advocating for and with survivors is helping assess needs and set goals to build a way forward. Autumn described the importance of advocates' skill sets including needs assessment and goal-setting: If your main goal is to get every woman here self-sufficient and learn about domestic violence so we don't get back in the situationy You take that person and you immediately figure out what they need, what they have. Take them and get stuff that they need so that the first priority can be, all right, you're ready to go look for a job.
Goal-setting can also help increase hope and facilitate empowerment. Leslie reported that in working with her advocate, Each day, we'd come up with resolutions. I would just go by each day and we would meet each morning and she just gave me so much insight that each day, everything became, every goal became closer.
Tammi described the empowering impact of strengths-based goal-setting in the wake of emotional abuse: They help focus you on certain goals that you need met and a lot of it really is empowering. If you've got somebody telling you forever you can't do anything or you don't do anything right or-a lot of people fall into that and when they're away from it, they don't know where to start from and it does give them a positive reinforcement. Yes, you're capable of doing this for yourself or getting employment to feel better about yourself, not just for money, but it helps redefine you or give you somewhere to start from.
Participants indicated that advocates must be ready and able to collaborate on what needs to be done, and when it needs to be done, in order to help meet the set goals. Sherri reported that it is helpful for an advocate to provide a step-by-step guide, "This is how this process is gonna go. You're here, you're trying to get here, but if you don't take these steps, you're not going to get [there]. It's laid out, simply." Sherri mentioned that, because "sometimes you're thinking about all this other stuff that's going on inside-junk inside of your head," she needed "somebody who can kinda keep me on the straight and narrow."

Discussion
For the last decade, research has indicated the need to make voluntary DV advocacy more survivor-centered (Davies & Lyon, 2014;Goodman & Epstein, 2008;Goodman et al., 2016c;Sullivan et al., 2017). However, there is a dearth of information from the perspective of survivors on how to shift advocacy practice to meet these goals. Interviews with 25 female-identified DV survivors in shelter and transitional housing helped to decrease this information gap. Participants emphasized the need for a safe space, both physically and emotionally, in order to build rapport and connection. It is critical that advocates make time and space to engage in listening to survivors' stories, and actively reflect that they have heard their needs. Further, advocates need to be accessible and available to survivors to facilitate connection and engagement. Feeling heard, as one survivor noted, opens hearts and minds for the work ahead. The perception of advocates as judgmental or unavailable can lead to service disengagement. Finally, advocates and survivors collaborate to meet their goals through education, resource connection, and goalsetting. These findings provide further support for Sullivan's (2018) assertion that fostering social and emotional well-being for clients with activities such as safety planning, skillbuilding, increasing resource access and social support, and empathy are critical for survivors.
There are several study implications to note. While safety is a bedrock of DV services, more attention is needed to the manner in which advocacy and other programmatic approaches hinder emotional and physical safety. This begins with intake and includes the built environment and agency rules (Wood et al., 2017). Refraining from judgment creates conditions where survivors can share their goals and needs without fear of repercussion. Questions and statements during hotline assessments, intake, and advocate-survivor interactions need to be phrased carefully to minimize the impression of judgment. Engaged listening helps promote what Goodman et al. (2016b) labeled alliance between the advocate and survivor. However, a lack of availability or access to the advocate can hinder alliance and reinforce feelings of isolation. The widespread implementation of voluntary service models promotes the empowerment goals of advocacy but requires continual engagement with survivors to make sure services fit needs (Nnawulezi et al., 2018). An implication of this study is the need to modify voluntary service practices to increase opportunities for survivor engagement while continuing to center agency and choice. This requires a careful balance of the survivor-driven orientation of voluntary services with the provision of support and information through regular communication, check-ins, and information sharing. Such voluntary service models can be challenging to implement because advocates may focus strongly on avoiding being "pushy" or presenting services as a mandate and end up being overly hands off, to the point that clients do not feel like they can reach out. For advocates, it can be labor-intensive to navigate this line effectively. Consistent with Weintraub and Goodman (2010), findings from this study support advocacy approaches that work both for and with survivors on their preferred goals.
The results of this study further suggest that a survivor-centered approach also reflects a trauma-informed care (TIC) approach. Participants in this study expressed the importance of and need for establishing safety and providing education about the impact of abuse reflecting components also aligned with TIC (Hedden, 2015). The TIC guiding principles as outlined by Hedden (2015) are safety, trustworthiness, peer support, collaboration and mutuality, empowerment, voice, choice, as well as attention to social location. The traumainformed model of voluntary advocacy, when applied to interpersonal violence, typically focuses on understanding the individualized nature of trauma and responding with services that prioritize safety and maximize opportunities to regain autonomy by offering survivors choice in service and support use and options (Nnawulezi et al., 2018;Sullivan et al., 2017). Trauma-informed models of advocacy have been previously noted to increase self-efficacy, feelings of safety, and empowerment (Sullivan et al., 2017). The findings of this study support the body of literature that suggests advocates should focus on resource acquisition for survivors, as well as providing information and referrals for a variety of individualized needs (Allen et al., 2004;Davies & Lyon, 2014;Sullivan, 2018). However, resource referral alone is not sufficient. These data suggest that while advocates should provide resources to their clients, they should also ensure that those are the best resources to meet survivors' individual needs and ultimately their goals. Furthermore, resources should be presented to survivors with a collaborative road map to help survivors understand what services may be useful, when, and in what order.
Working from a survivor-centered, trauma-informed perspective requires a balance of engagement and listening that takes patience, skill, and emotional energy, in addition to organizational support (Nnawulezi et al., 2018). As DV advocates already face significant challenges with low pay and occupational stress (Kulkarni et al., 2013;Slattery & Goodman, 2009), using a trauma-informed and survivor-centered model also necessitates organizational support and ongoing training to help advocates thrive in their roles. As suggested by previous research, this may include increasing advocacy staff, providing high-quality supervision, and peer support (Kulkarni et al., 2013;Slattery & Goodman, 2009). Further, additional advocacy skills may be needed to best implement survivorcentered approaches and balance advocate workload, such as the use of social networkoriented approaches facilitated by peer engagement (Goodman et al., 2016a) which aim to reduce isolation and increase support from the community. Social network approaches may be especially important for helping survivors create new social networks and to reengage with old networks for increased support (Goodman et al., 2016a).

Limitations
While this study offers critical survivor insights into the continued development and improvement of survivor-centered models of advocacy, limitations and the need for further research remain. Given the scope and exploratory nature of the study, we did not investigate potential differences between the four agencies based on program structure, advocate training, and guidelines for service, leading to an inability to assess one advocacy model in particular. Additionally, this study does not examine differences in advocacy needs based on factors such as race, ethnicity, gender, class, religion, immigration, or disability status, which may be critical to understanding survivor perspectives. Future research should use a larger and more diverse sample and begin with an intersectional design to examine these critical issues. Furthermore, while DV continues to disproportionately impact women, the authors acknowledge that a wide range of individuals, including men and those who identify as gender-queer and non-binary experience DV. The field will benefit from future research that explores the role of gender, as well as additional factors that may impact the needs of survivors and advocacy approaches in order to ensure they are appropriate and effective. Furthermore, this study was strictly qualitative and focused on perceptions and experiences of advocacy. Future studies incorporating quantitative methods may further examine implementation factors of advocacy models such as feasibility and resource use.

Recommendations
This work points to the need for more research on survivor-centered, trauma-informed advocacy model implementation and fidelity, expanding the work of previous studies on the CAP, VDA, and other models of survivor-centered practice (Allen et al., 2004;Bybee & Sullivan, 2001;Goodman & Epstein, 2008;Goodman et al., 2016c;Sullivan & Bybee, 1999). The findings from this and other studies need to be tested on a larger scale with mixed-methods assessments. Evaluation is needed to understand the outcomes of different survivor-centered advocacy approaches and how they compare with other service provision models. Outcome evaluation work is also needed to assess the efficacy of differing models, particularly considering issues around mental health, addiction, and cultural intersections and adaption. Additionally, more information is needed to guide the use of voluntary service models that typically accompany a trauma-informed, survivor-centered approach. The small body of literature on voluntary service models suggests it is an empowering approach (Nnawulezi et al., 2018), but there is a dearth of information to guide implementation. Without this evidence, data suggest that implementation will be less consistent (Damschroder et al., 2009). The domestic violence field would benefit from future research and evaluation to establish models of voluntary service advocacy using survivor-centered and trauma-informed approaches.

Conclusion
Services for DV survivors have long used empowerment and agency-building strategies in advocating for survivors' needs and rights. However, changes in DV services have shifted the ways in which advocacy operates and is often defined by service providers rather than survivors. Using survivor-centered approaches may help to enhance the traumainformed model in DV services, and support a voluntary service model. This study further highlights and delineates the needs and corresponding skills survivors may need from their advocates. Increased understanding of survivors' advocacy needs, approaches to advocacy services, and advocacy skills required for service will lead to improvement in the overall quality and effectiveness of care for DV survivors. The findings of this study present a strong foundation from which future research can further explore the best practices for core DV advocacy services.