Gendered Practices in Child Protection: Shifting Mother Accountability and Father Invisibility in Situations of Domestic Violence

This article reports on an exploratory, qualitative, multiple-methods study that included individual interviews and a focus group with child protection services (CPS) workers in a large city in Alberta, Canada. The findings illuminate current CPS worker practices in situations of domestic violence where inclusion and exclusion decisions are made for service provision, and the ways in which documents reflect these day-to-day practices; how service user descriptions are constructed and reconstructed, the social problem of domestic violence conceptualized, and the ways in which professional development training encourages critical thinking about existing practices to create new solutions for families experiencing domestic violence. Thematic analysis reveals three themes about CPS workers’ experience: 1) current practices reflect invisibility of men and accountability of women; 2) personal and professional shift in perspectives on who to work with, gender expectations, and how CPS are delivered; and 3) reflexive practice into potential intervention strategies and professional development training. The findings suggest specific recommendations for practice including the need to engage men in child welfare practice, shift perspective about service delivery with families experiencing domestic violence, and account for gender norms and practices in service delivery.


Introduction
This article presents a described shift in child welfare practices as a result of professional development training and the response of child protection services (CPS) workers to the training. Prior research recommends the shifting of child protection practices in situations of domestic violence to enable the support of mothers (survivors) while holding fathers (perpetrators) accountable (Hughes, Chau, & Vokrri, 2015). Our study, situated in Alberta, Canada, adds an important contribution to the research literature specific to changing practices in child protection.
The prevalence of families who are affected by domestic violence in Alberta is high, growing along with the number of child witnesses to domestic violence (Government of Alberta, 2014). Alberta has seen a 2% increase since 2014, ranking third highest out of all Canadian provinces for domestic violence (Statistics Canada, 2015). Similarly, according to the 2008 Canadian Incidence Study of Child Abuse and Neglect (CIS), one of the most frequently occurring categories of substantiated cases of child maltreatment (34% or 29,259 cases) was exposure to intimate partner violence (Black, Trocmé, Fallon, & MacLaurin, 2008).
Researchers note that professionals providing services and support to families who have experienced do-mestic violence often lack training around domestic violence (Fotheringham, Dunbar, & Hensley, 2013;Hughes et al., 2015). In Alberta, there are multiple influences that guide child protection policies and practices when engaging with families experiencing domestic violence. For example, risk assessments are guided by the "best interests of the child" standard, and domestic violence is recognized as one factor while, at the same time, the 2003 Provincial Family Law Act maintains a presumption after marital breakdown that both parents are guardians of their children (Boyd & Bertrand, 2016). Indeed, a study of legal professionals' perceptions of shared parenting reveals an understanding that shared parenting rates are higher in Alberta than in other parts of Canada (Boyd & Bertrand, 2016). In neighboring British Columbia, child protection policy acknowledges the need to keep mothers safe and support her in the care of her children (BC Ministry of Children and Family Development, 2017).
Drawing on the White Ribbon Campaign's issue brief Engaging Men and Boys to Reduce and Prevent Gender-Based Violence (Minerson, Carolo, Dinner, & Jones, 2011), and the findings from Alberta-based Shift: The Project to End Domestic Violence (Wells et al., 2013), a local collective of service providers and academics created curriculum training for the community that focused on examining male normative ideas (male norms) about masculinity as a means of violence prevention. The curriculum workshop was piloted with local professionals working in the fields of sexual and domestic violence. A local child protection organization in Alberta, Canada, requested the workshop to facilitate their understanding of the link between masculinity and domestic violence. In our article, we explore CPS workers' perceptions and current practices with families experiencing domestic violence, and CPS workers' understanding of the link between domestic violence and masculinity. Our findings illuminate CPS workers' reconstructions of service users and child welfare practices. Specifically, CPS workers indicate that they are now more critical of the parent role, moving beyond mothers as the sole responsible parent, and they make further attempts to engage fathers in their services. Additionally, CPS workers state that they no longer write statements in the agency files to implicate mothers as solely responsible for child protection, and they include information about the father in the agency data base.

Literature Review
Domestic violence is a major social problem in Canada. When children were present in the home during violent incidents, 59% of women reported that their children heard or saw the violent act (Statistics Canada, 2013). Within a context of violence, mothers are held to a higher standard than fathers in protecting their children (Boyd, 2017;Hughes et al., 2015). Hughes et al. (2015) point out that women experiencing violence shoulder inappropriate blame from CPS workers for the impact of the vio-lence on their children; the systems that are designed to protect mothers, instead, construct her as an unfit parent. Additionally, the family law system may judge mothers more harshly for not protecting children from violence, and CPS may question the mother's motivation to keep their child safe (Boyd, 2017;Hughes et al., 2015). Jevne and Andenaes (2015) highlight this same finding in a study of 15 parents, where two mothers expressed safety concerns to professionals, leading to the loss of maternal custody, with limited supervised access to her children. In a study examining how Family Courts remove children from their parents, Mosoff, Grant, Boyd and Lindy (2017) suggest there are numerous such examples of mothers losing custody of her children to the state in our Canadian child welfare system. The authors found that CPS removed children from their mothers in situations when the father or male figure in the home created a risk to the children through violence or criminality (Mosoff et al., 2017), representing the gender bias that Bancroft and Silverman (2002) describe in the United States.
Women endure scrutiny for their inability to protect their children in situations of domestic violence while fathers are under-involved in the process of keeping their children safe (Alaggia, Gadalla, Shlonsky, Jenney, & Daciuk, 2015;Humphreys & Absler, 2011). According to Alaggia et al. (2015), 63% of perpetrating parents (predominantly fathers) were unreachable during child welfare investigations; while the survivor of violence (predominantly mothers) were investigated in over 90% of all situations. These statistics translate into potential practice of convenience (CPS workers engage with mothers only) rendering the perpetrator invisible.
The Government of Canada's report on child abuse and neglect found that domestic violence was present in 34% of substantiated child welfare investigations (Black et al., 2008), while child intervention staff in Alberta continue to receive less than five hours of family violencerelated training (Snyder & Babins-Wagner, 2012). This is particularly relevant, considering that exposure to appropriate training tends to lower workers' negative views of families experiencing domestic violence, increase empathy with survivors, and increase workers' willingness to engage perpetrators, holding them responsible for their actions (Snyder & Babins-Wagner, 2012).
In Alberta, changes to child protection and family laws have recently tried to shift current "mother blaming" practices (Humphreys & Absler, 2011). In child welfare there has been provincial adoption of "Signs of Safety", an internationally recognized strengths-based, safety-oriented approach to family casework. Also impacting CPS practices is a change in family law adding exposure to domestic violence as a mandatory reporting requirement and a factor within the "best interest of the child" standard (Cross, Mathews, Tonmyr, Scott, & Ouimet, 2012;Family Law Act, 2003). The legal system increasingly recognizes that physical violence and coercive control are important components of domestic violence; this shift is significant because family court has historically minimized or denied the destructive presence of coercive control, usually of men over women, through physical intimidation, social isolation, withholding access to finances or resources, or through seeking partial custody of the child (Elizabeth, 2015).
Despite these changes, some research indicates that CPS continue to hold mothers to higher parenting standards than fathers (Hughes et al., 2015;Humphreys & Absler, 2011). Following incidents of domestic violence, mothers become the focus of child welfare investigations with mothers reporting that they feel a lack of support in addressing the problems of domestic violence (Hughes et al., 2015). The courts maintain different parenting expectations for mothers and fathers in child custody decisions; the family court system often weighs positively fathers' expressions of caring about their children, while taking for granted the day-to-day labour involved in caring for the children, a task still overwhelmingly performed by mothers (Boyd, 2013). Further, family law courts routinely order father access to children despite male violence against women and children (Boyd, 2013), while mothers can be perceived to have failed at protecting her children in situations of domestic violence despite the absence of her abusive behaviour (Mosoff et al., 2017).
Research about frontline CPS workers' experiences and their understandings of the linkages between masculinity and domestic violence is limited (Wells et al., 2015). Shift: The Project to End Domestic Violence, an Alberta-based research project, reveals: 1) only a small number of programs are focused on domestic violence prevention or advancing gender equality; and, 2) a link exists between current norms of masculinity and domestic violence (Wells et al., 2015). Much of the province's current domestic violence programming is focused on crisis response and victim services, though some programs have emerged to support fathers, such as the Alberta Father Involvement Initiative, and the province has seen a growing number of school-based initiatives promoting healthy masculinities, relationship skills and gender equality among boys (Wells et al., 2015).
The lack of information on CPS workers' understanding of the link between masculinity and domestic violence represents a significant gap in the academic and practice research with potential implications for the ways in which CPS workers interact with fathers and mothers. In naming parents as either mothers or fathers we are not intending to ignore same sex or nonbinary-identified parents, however in this study, CPS workers spoke of a parent gender binary. Some researchers claim that explicitly highlighting the role of gender in child protection may lead to a greater understanding of CPS workers' challenges engaging fathers in discussions about their children's safety (Baum, 2015;Scourfield, Smail, & Butler, 2015). Recent changes in British Columbia reflect the inclusion of a family development response (FDR) in child protection when domestic violence is present wherein mothers, fathers and children are included in safety assessment and planning. However, child-centered and mother-centered approaches appropriately trump FDR when the safety of the mother and children remains a concern, therefore, these families are not expected to participate in couple counselling, family mediation, anger management, or visitation arrangements (BC Ministry of Children and Family Development, 2017).
Our study examined the perceptions of CPS workers after their participation in professional development that consisted of a full day of curriculum training (lectures, videos, small and large group discussions) inviting reflection on personal perceptions of male violence, masculinity, and gender roles in relation to child protection practices. Specifically, we wanted to explore two questions. First, how do CPS workers understand current practices with families experiencing domestic violence in the ways they construct mothers and fathers? Second, in what ways does professional development training specific to the link between domestic violence and masculinity support CPS workers to shift their practices with families and how they record family information? Ethics approval for the study was received from the Internal Review Board.

Study Design
This study uses a qualitative research design, seeking information directly from participants about their experiences (Denzin & Lincoln, 1994. This methodology allowed for a greater insight into the child protection worker experience of the curriculum workshop and their understanding of the link between masculinity and domestic violence. Multiple methods were used in this study including two rounds of individual interviews and a focus group. Following the interviews, we engaged participants in a focus group to gather qualitative data during a two-hour session with a homogenous group; child protection workers (Krueger & Casey, 2015). The focus group provided greater clarity of the interview data as participants shared further insights and connections (Krueger & Casey, 2015).
The research team consisted of the primary investigator (PI) and two graduate research assistants (RAs). We received permission from the Alberta government to recruit CPS workers for this study following a request from a local child welfare agency seeking professional development curriculum training for their CPS workers that would focus on examining the link between male norms and domestic violence. The professional development workshop was facilitated by one of the creators of the curriculum training, a local professional and member of the Calgary Domestic Violence Collective and subcommittee member of Engaging Men and Boys (EM&Bs). Twelve CPS workers from a local child welfare agency in a large city in Alberta, Canada, engaged in the full-day professional development curriculum training workshop. Utilizing purposive sampling to allow all participants opportunity to answer the research questions (Patton, 2002), the PI and one RA recruited study participants from the twelve attending CPS workers, taking time at the beginning of the professional development session to share an overview of the study and leaving behind the RA contact information for potential participants. Participants were invited to be part of the study based on their interest in sharing their work experiences in domestic violence and their willingness to contribute to curriculum development focused on male norms and violence prevention. Interested participants were invited to provide contact information for a follow-up interview. Nine participants initially indicated an interest in the study, while seven participants followed through with two interviews.
The CPS workers belonged to units within the organization that oversee families involved in domestic violence. Participants were university educated with a bachelor's or master's degree in a range of disciplines such as social work, psychology, counselling psychology, and child studies. Participants had between three and sixteen years of experience working in child welfare, and their ages ranged from 30 to 58 years old. Six of the participants self-identified as female, and one male, while all identified as Caucasian.
Data collection included one-on-one interviews (1.5 hours long) conducted by a RA with each participant sharing their perceptions and experience about the linkages between norms of masculinity and domestic violence, including what they learned in the focused training, and their understanding of their perceived training needs. There were six interview questions, each with further prompts. For example, interview question one was about participants' experience with the EM&Bs' training, followed by four prompts including, anything new that they didn't know, anything surprising, anything they didn't agree with, and how they understood the link between domestic violence and masculinity. Question two asked about the ways in which they have been influenced by the training including a shift in their thinking. Question three explored the ways in which the training will influence their future practice with perpetrators, survivors and children. The remaining questions focused on other professionals who might benefit from this professional development and future professional development they would like to see incorporated in their workplace. A second round of individual interviews was held to further explore participant descriptions and meanings of their experiences. All interviews were audiotaped and transcribed verbatim.
Based on the findings from the initial analysis of the interview data, the research team drafted an interview guide for a focus group, also transcribed verbatim. Of the seven study participants, five agreed to engage in the focus group. The two-hour focus group, facilitated by the PI and one RA, provided a means of member check and ad-ditional data collection, such that the focus group members reflected on the initial emergent themes in terms of how the data resonated or did not resonate with their understandings and experiences. Analysis of the interview data and focus group data allowed for triangulation and increased trustworthiness of the study findings.

Data Analysis
Thematic analysis was used to analyze the data collected (Clarke & Braun, 2017). While recruitment for the initial interviews took place on the day of the curriculum workshop, recruitment for the focus group, data collection and data analysis were an iterative process (Braun & Clarke, 2006). After the interviews, the research team reviewed the transcripts multiple times. During analysis, the data was manually organized into codes, categories, themes and sub-themes to identify patterns in the data (Clarke & Braun, 2017). Two members of the research team, the PI and one RA, independently developed codes from the raw data using a line-by-line approach (Braun & Clarke, 2006). A table was utilized to capture raw data, codes, categories and themes for ease of organization and review. Categories were created to organize the codes and reviewed for consistency. Together, two research members reviewed the codes and categories to create the initial themes and subsequent subthemes that were shared with study participants during the focus group. Thematic analysis was used to analyze focus group data, following a similar process described in the interview data analysis.

Findings
The researchers identified three overarching themes which were illustrated in multiple ways by various study participants. Themes were further reduced to include sub-themes (Braun & Clarke, 2006). The analysis describes existing everyday processes of parental inclusion and exclusion in the child welfare system. Themes are illustrated through participant quotes.

Theme 1: Current Child Protection Practices Reflect the Invisibility of Men and Accountability of Women
This theme represented the ways in which participants talked about current practices including what occurred day to day in their work, areas of practice that went unchallenged and where they believed they lacked critical awareness, gaps in service regarding male inclusion in accountability, and unexamined biases regarding the role of mothers. Many participants expressed that current professional development practices excluded domestic violence. They also noted that current CPS worker practices in domestic violence cases revealed a bias toward mother accountability and father invisibility, which they explained became clear to the workers themselves during the professional development day. For example, par-ticipants described a number of myths that drove their practice, which appeared to influence their day-to-day practice decisions, as indicated in the following quote: We are feeding into those myths, that it's the woman who is in charge to ensure the safety for the children and we just sort of let men off and it is just sort of that myth, you know, that they [fathers] are there to be the bread winners.
Fear, safety, and relationships were some of the reasons that participants provided for their decision to work exclusively with mothers. For example, CPS workers feared working with fathers for their own safety and feared destroying relationships with mothers. One participant suggested: "Sometimes it is the risk thing…they are, like, well, he is violent, we can't go to the home". Another participant indicated: "It's more an alignment, like, we work with [the] mom…we don't want to risk that relationship if we work with [the] dad".
As a result of the myths, fears and concerns for safety, CPS workers developed constructions in child welfare practices presented next as a sub-theme.

Constructions in Child Welfare Practices
Participants identified constructions in daily practice such as parenting responsibilities and included who to engage, responsibilities for child protection, and CPS workers' role. For example, parental responsibility focused on the parent who could be engaged and those who cannot be engaged were excluded. Participants noted they avoided father engagement because of a lack of response, as noted here: You get a family violence file, and then you work on calling [the] mom, and you do safety planning with her, and you can't get a hold of the dad and you just go, well, I tried and he wouldn't engage, and then you let it go.
The role and responsibility of protecting children was left to the parent more easily accessible, often the mother who was caring for the children. The CPS worker then viewed their role and the responsibility inherent with a mandate of protecting children, as the enforcer; the CPS workers enforced the role of mother as protector as identified here: In child intervention services, I think we still largely focus on the women's role in it…it is around protecting her child and leaving her partner and, like, enforcing an emergency protection order, or a restraining order or whatever kind of, like, legal sanction she has, like, being the enforcer of it.
In their role of enforcer, CPS workers held mothers to be the one who takes responsibility. In the case discussed here that may imply that they use the administrative route, acquiring the correct document and following the legislation. They explained that they used to expect: [Mom to] go to court and get a restraining order and follow through with the terms and conditions. And our legislation, our legal authority on that child will correspond with your ability on following through with the protection order that you have.
Indeed, participants asserted the commonly held belief that service was no longer required when the mother was protecting the children and documentation in the file reflected this practice; for example, "I look back at [the] history and what happened….The file is closed because mom is protecting, that would be common language".

Theme 2: Personal and Professional Shifts in Perspective About Who to Work with, Gender Expectations, and How CPS Services Are Delivered
Following the professional development day, the CPS workers shared some of the shifts that had occurred in their thinking about their practice and the way they went about their work. For example, they said that they contact fathers and schedule meetings with him, they have discussion in their team meetings about their meetings with fathers, and they include information about fathers in the agency data base. They discussed the need to engage men in their practice, to shift their perspective about what child protection meant, and to account for gender norms and expectations in their understanding of service delivery. CPS workers began to reconstruct the meaning of child protection as participants noted the importance of critically examining their use of language and how this shifted for them following the training. For example, participants shared how their perspectives and previously held knowledge were challenged: But I never just thought about, like, men, like, that being the issue. About how they were raised, or their beliefs, or what they think about masculinity, and what it means to them. And I never thought about that, so it [professional development] kind of challenged that.
Participants agreed that the training helped them to deepen their understanding of the link between masculinity and domestic violence and shifted their perspectives as presented in the following sub-theme.

Developing Reconstructions of Child Welfare Practices
The shift in perspective for CPS workers began a dialogue about reconstructing child protection. Participants' greater understanding was evident when they discussed the need to include fathers in conversations about domestic violence.
Just asking, have you talked to dad, does dad have a network, and then I have actually gone with workers now who are starting to have meetings with dad and having those conversations started that maybe wouldn't have [happened] before.
CPS workers indicate that they are having in person meetings with fathers at the office, inviting them in for discussions and constructing him as a parent with responsibility for child protection. This new understanding has participants rethinking domestic violence, and the impacts of how service might be provided as noted by one participant: "What I am finding now, you know, if you have a little more empathy and kind of look at things a little deeper, then you might actually make some more lasting change". And another participant here as they considered including the father in the role of parent with responsibility for child protection: I don't know if it's that we think that the perpetrator can't change or that it's too much work to ask them to change or what that is, but I know we do that all the time. So, I think we need to change that in our practice altogether.

Theme 3: Reflexive Practice into Potential Intervention Strategies and Professional Development Training
The professional development day when CPS workers gathered together with colleagues and the facilitator raised for them significant areas requiring further development. This included the need for more reflexive practice around constructions of masculinity, privilege, and power that influenced their interactions with clients. Reflexive practice during and after the day's session highlighted the link between masculinity and domestic violence, opened space for potential new practices and intervention strategies to emerge, and pointed to the need for additional professional development. Following the professional development training, CPS workers discussed practices (meeting with fathers, including fathers in the agency data base, relying less on mothers to be the sole protector of children) where the parents were reconstructed as fathers and mothers, both included in service provision and both responsible for child protection. Participants indicated that reflexive practice provided direction for intentional practice: I was writing more about what dad said and what mom said and what the plan was and is, and that is why we can close the file. So, making sure that dad is involved to talk to, so I didn't let that go anymore. And making sure I had a face to face with them, and then making sure I had a conversation about um, how he was raised and his family.
Taking the step to have meetings with fathers meant that CPS workers were gathering more and different in-formation about the family and about child protection. This information was being recorded digitally within the agency data base, supporting the notion that both parents are responsible for child protection. Service plans included fathers' role and responsibilities. Mothers were not identified as the only parent in the agency file with sole responsibility for child protection because fathers were included in discussions about their responsibility for child protection. While reflecting on their (lack of) practice with men CPS workers developed new constructions in child welfare practices, as presented in the following sub-theme.

Reflexive Practice Supports New Constructions in Child Welfare Practices
Participants suggested in the interviews and focus group that professional development helped them be more reflective about language and also shifted their language. This occurred for them in their daily practice interacting with colleagues and service users, and how they reported and documented in the agency files. Here the participant revealed their understanding of dangerous practices that left mothers solely accountable for the protection of children, and how their increasing reflexiveness resulted in changing practices: "I am not writing those statements anymore, 'mom is protecting'. Like, I am not doing that anymore".
Here, mothers were reconstructed as not solely responsible for child protection, and the agency file was rewritten with a different construction of mothers. CPS workers described one way in which reflexive practice assisted in their daily practice providing support to their peers during team meetings: [During part of the training] she changed the wording because it sounds a little bit like victim blaming. And I know that my co-worker was emotionally charged about that particular instance, and so she was kind of mad at that victim and I think that happens, so holding each other accountable.

Discussion
The findings from this study begin to fill the gap in our understanding of the ways in which CPS workers interact with mothers and fathers in situations of domestic violence and challenges the existing welfare rhetoric of inclusionary practices in domestic violence (both parents have access rights to their children). The study findings also highlight the ways the existing child welfare practices can reinforce harmful practices of control. Participants suggest that men have historically been excluded from the role of responsible parent, by CPS workers under-involving him in the plan to protect his children, and by eliminating his role as protector in the agency file. In this way, participants indicate that violent men have not been held accountable as responsible parents.
Instead full expectations and responsibilities have been placed on mothers for children's safety. These findings are consistent with recent studies involving parents that reveal mothers are held to a higher standard than are fathers (Boyd, 2017;Hughes et al., 2015) and fathers are under-involved in child protection (Alaggia et al., 2015). The findings are now also clear from the perspective of CPS workers. Luther (2015, p. 16) argues that one of the historical purposes of child welfare legislation has been to address "the problem of intervening in families of poverty". Further the author states that throughout the history of child welfare, many professionals (medical, legal) have had influence over what constitutes harm to children, developing constructions of mothers and children types that are then used to identify who requires intervention; reinforcing relations of power for marginalized families (Luther, 2015). The application of concepts such as deservedness, harm and best interest evaluations on poor families within a neoliberal environment invite a "mother" focus for intervention given the presumed social obligation of child caregiver (Luther, 2015). Indeed, in her dissertation, the author poses for consideration the ways in which our Canadian dual legal system (those with means, those without means) is reinforced by our child welfare policies when considering rights and duties. For example, we support fathers' (those with means) rights to mother and child access, yet we do not enforce their duty to be a protective parent. Similarly, we do not support mothers' and children's (those without) rights to safety, and we expect mothers to solely fulfill her duty as a protective parent.
Our study suggests that current child protection practices reflect tensions between provincial child custody legislation (both parents gain custody and access) and child protection policies (skewed toward mother accountability) in situations of domestic violence. Highlighted in participant quotes (theme one) participants illustrate awareness of their current practices, including the way child protection practice standards generally support provincial child custody legislation (Family Law Act, 2003) where both parents gain regular and meaningful contact with children, while in daily child welfare practice only the mother is held responsible for her children's safety. Participant statements suggest a new awareness about the ways their practices, holding mothers accountable while granting fathers access to the children, place mothers and children at greater risk of harm. The implied message to the mother is that the system will support her if she follows through on certain terms and conditions (e.g., getting a restraining order), while the father holds no commensurate expectation or accountability for protecting his child from harm (Snyder & Babins-Wagner, 2012). The professional development appears to have increased CPS worker awareness of the complexities inherent in shared parenting specific to situations of domestic violence. For example, theme two reveals increased awareness of the need for domestic violence training, while theme one illuminates gender expectations that can leave mothers and children at risk of harm. Further, participant responses highlight current limited domestic violence professional development for CPS workers, a lack of awareness about how best to change this dynamic and, a gap in knowledge of how to better engage fathers in the child protection process. Baum (2015) suggests that utilizing a gender lens in situations of domestic violence supports CPS workers to recognize power relations, mutual fear, and communication differences that exist in CPS worker/father relationships, sub-themes arising in this study. For example, participants suggest a shift in understanding away from the current focus on mothers as sole protectors of children to a greater focus on the father's responsibility for child safety and even the responsibility of CPS workers to support children's safety. Three examples are provided in the study data: 1) reflecting on gender norm expectations in theme two; 2) identifying power relations and mutual fear in theme one; and 3) reflective and intentional practice in theme three. While within a patriarchal society, socially men hold more power than women, within the CPS worker-father relationship, some believe that fathers may think that the typically female CPS worker holds power over him and poses a threat to the integrity of his family unit (BC Ministry of Children and Family Development, 2017). Others argue the father may feel threatened or vulnerable disclosing potential feelings of inadequacy as a parent, leaving him to act aggressively towards the worker (Baum, 2015;Scourfield et al., 2015). In turn, the worker may fear interaction with him; a fear that may increase should there be a shift in practice toward engaging fathers who have used violence with their partners. Some authors suggest that the fear dynamic may create a difficult work environment for establishing trust between service provider and service user in situations of domestic violence when engaging men in CPS work (Baum, 2015;Scourfield et al., 2015).
Engaging men in CPS practices in situations of domestic violence is not common according to participants in this study, yet the professional development training supported CPS worker reflexive practice, suggesting a greater engagement of father in child protection and documentation of fathers in agency files. Some participants indicate that fear is at least part of the reason why engagement of fathers is not pursued. The fears reported by participants in this study about working with men who have been violent with their partner are real. Safety planning for CPS workers may be necessary when practices shift to include further engagement of men who have been violent with their partner. CPS workers will need an opportunity to discuss their fears and create plans of engagement. These might include worker safety such as meetings in public spaces or at the CPS office, and the inclusion of a support person such as a colleague in meetings (BC Ministry of Children and Family Development, 2017).
In theme two, participants note a shift in perspective on gender roles and expectations; a shift on a professional and personal level. This shift has caused them to critically evaluate their practices of working with mothers. These practices are reflected in the research literature suggestive of a gender bias in CPS practices (Bancroft & Silverman, 2002). Participants' deeper awareness highlights how reflexive practice can help CPS consider the implications of their daily work, and reconsider alternate ways for inclusive practice. For example, participants note that they have begun to question what including fathers in child protection means. Their thinking about who is responsible for ensuring the safety of children has evolved away from an exclusive focus on mothers as the responsible parent. A deeper gender analysis of the CPS worker-service user relationship can further our understanding of this dynamic.
Participant responses suggest strategies that CPS workers can use in their day-to-day practice to shift the culture towards engaging fathers in child protection. Some of the emerging strategies under theme three included: holding each other accountable, being more mindful of power and language, and incorporating reflexive practices around engaging or not engaging men.

Limitations
The study findings reflect the views of a limited number of CPS workers from one organization in Alberta, Canada, and cannot be assumed to represent the experiences of all CPS workers. Additionally, given the qualitative nature of the study, the findings are not generalizable. Participants self-selected to be part of the study and may have a greater interest in this topic than other CPS workers within their organization. This study focuses exclusively on domestic violence in heterosexual relationships from a binary understanding of gender; it does not address domestic violence situations involving same sex or nontraditional relationships.

Implications
The 2008 Canadian Child Incident Study found that the single greatest risk factor for child maltreatment was poverty, that many families who come into contact with CPS workers are poor, and that the number one safety risk factor for the primary caregiver is domestic violence. Together, these findings suggest a really important message for our child welfare policies and practices about the types of services that these families require. Instead of the traditional services offered by CPS workers to families, such as parenting education, the focus needs to shift towards equipping women and children who have experienced violence, with critical supports to address their poverty-related challenges (Luther, 2015). Currently, the child welfare system operates with risk assessments and utilizes coercive interventions (removal of children from parent) to legitimize marginalized women who do not fit into social norms, centering practices on what CPS workers identify as mother deficits (Luther, 2015).
Participants in this study say that CPS workers would benefit from training designed to highlight gender roles and expectations in situations of domestic violence. Such training may support the workers in their difficult day-today decisions about who to engage in the protection of children, and in reconstructing parents' roles and responsibilities for the safety and protection of children. Fineman's (1999) work suggests that child welfare policies need to move away from a system of coercive state intervention so that women and children can be supported through a collective responsibility. This would mean that people beyond mothers, such as CPS workers, fathers, and government, would hold responsibility for child and mother safety and well-being in situations of domestic violence. The findings from this study suggest that the CPS workers who participated in the professional development training have developed greater awareness about the implications of current practices that hold mothers solely responsible for child safety; participants show signs of shifting their perspective through reflexive practice. Indeed, the professional development training supported workers to reconstruct their understanding of service users and of child welfare practices. Fineman (1999) tells us that in order to shift our view of mothers as a risk to their children, we need a more realistic understanding of the associated challenges in situations of domestic violence such as poverty. Gaining an understanding of domestic violence and its link to masculinity is an important step in supporting CPS workers in their day-to-day work with families to keep mothers and children safe. Inviting fathers into the conversation about their role and responsibility is to reconstruct them as parent with obligations for the protection of their children. For CPS workers it means arranging meetings in safe spaces such as the office or other public space when safety is a concern. CPS workers will need to document roles and responsibilities of fathers, mothers and CPS workers in service plans and agency data bases, reconstructing our understanding of what it means to protect children.