Evaluation of cross-disciplinary training on the co-occurrence of domestic violence and child victimization: overcoming barriers to collaboration.
Abstract: The co-occurrence of child abuse and domestic violence has gained increasing attention over the last several years. As a result, there have been a number of efforts around the country to cross-train domestic violence and child welfare workers. Many of these initiatives are based on recommendations derived from prior research which emphasizes the importance of cross training child welfare workers, domestic violence advocates, and others in order to enhance interagency collaborations and ultimately improve the handling of co-occurrence cases. Using a survey design and two samples of child protective service (CPS) workers were drawn, this paper evaluates the effectiveness of a statewide initiative to improve inter-agency collaboration through a series of cross-disciplinary trainings. Special attention is also given to the identification of barriers and their potential role in shaping collaboration. While this study did not find significant changes in CPS workers' knowledge, attitudes, or self-reported levels of collaboration overall as a result of the training, collaboration was found to be related to increased knowledge and positive attitudes toward collaboration post training. The findings further demonstrate that efforts to cross-train staff can change the way CPS workers view the presence of some barriers to collaboration. Implications for future research and strategies for enhancing inter-agency collaboration in co-occurrence cases are discussed.

Keywords: collaboration, co-occurrence, cross disciplinary training, barriers, domestic violence, child victimization
Article Type: Report
Subject: Child welfare workers (Training)
Child abuse (Control)
Family violence (Control)
Child welfare (Human resource management)
Authors: Haas, Stephen M.
Bauer-Leffler, Simon
Turley, Erica
Pub Date: 12/22/2011
Publication: Name: Journal of Health and Human Services Administration Publisher: Southern Public Administration Education Foundation, Inc. Audience: Academic Format: Magazine/Journal Subject: Government; Health Copyright: COPYRIGHT 2011 Southern Public Administration Education Foundation, Inc. ISSN: 1079-3739
Issue: Date: Winter, 2011 Source Volume: 34 Source Issue: 3
Topic: Event Code: 280 Personnel administration Computer Subject: Company personnel management
Product: Product Code: 9101226 Domestic Violence (Families); 9101224 Child Abuse NAICS Code: 92219 Other Justice, Public Order, and Safety Activities
Geographic: Geographic Scope: United States Geographic Code: 1USA United States
Accession Number: 304051559
Full Text: INTRODUCTION

Over the last thirty years, researchers have studied the association between domestic violence and child victimization and have confirmed that such behaviors often occur within the same family. In fact, many studies have concluded that the co-occurrence of domestic violence and child victimization is rather substantial and wide-ranging (Appel and Holden, 1998; Edleson, 1999). In a review of thirty-five studies published between the late 1970's through the 1990s, it is estimated that adult domestic violence and child maltreatment was co-occurring in 30 to 60 percent of families (Edleson, 1999; Banks, Hazen, Coben, Wang, & Griffith., 2008; Potito, Day, Carson, & O'Leary, 2009). More recently, reviews of child protective services cases have disclosed that roughly 4 in 10 cases involving children who were either killed or seriously injured, the children had also been exposed to other domestic forms of violence (Spears, 2000). There is also some evidence that the severity of violence exhibited among intimate partners is closely associated with the severity of abuse experienced by children living in the home (Appel and Appel, 2006). Yet, it has only been within the past ten years that major efforts have been undertaken to address the co-occurrence of domestic violence and child victimization.

Historically, these two forms of victimizations have been treated independently of one another--even when members of the same family are involved. This is in part due to the fact that the two systems in charge of dealing with such victimization have operated under different sets of philosophies, missions, mandates, policies, and procedures (Appel and Appel, 2006; Moles, 2008; Potito, et al., 2009). While domestic violence programs more often deal with violence between adults, child protective service agencies are responsible for supporting, defending, and advocating for child victims of abuse and neglect (Saunders and Anderson, 2000; Banks, et al., 2008). Likewise, child welfare agencies often reside within state and other governmental entities while domestic violence service programs are typically housed in nonprofit organizations. As a consequence, the relationship between these two systems has sometimes been characterized as one with a lack of trust and where cooperation is more of the exception rather than rule (Findlater and Kelly, 1999a; Potito, et al., 2009).

Nonetheless, communities have begun to recognize that one agency alone cannot properly address the needs of families experiencing both domestic violence and child victimization. In response, new strategies centered on the use of cross-disciplinary training to increase collaboration are developing across the country on both state and local levels (Nuszkowski, Coben, Kelleher, Goldcamp, Hazen, and Connelly, 2007; Banks, et al., 2008). For example, states such as Michigan and Massachusetts are leading the way by successfully conducting cross-disciplinary, co-trainings among domestic violence and child welfare workers in an effort to improve the handling of co-occurrence cases (Whitney and Davis, 1999; Saunders and Anderson, 2000). Los Angeles and Orange counties in California have also sought to change the attitudes and knowledge of personnel across disciplines in an effort to increase inter-agency collaboration (Mills and Yoshihama, 2002).

Such initiatives are typically accomplished through didactic methods of instruction and the delivery of specific information designed to foster a better understanding of inter-agency policies, procedures, and responsibilities. Moreover, such trainings often also seek to educate trainees on how to recognize and overcome barriers to collaboration. By changing individual attitudes, educating case managers on the roles and responsibilities of other system actors, and facilitating change in prevailing policies and procedures that inhibit inter-agency cooperation, it is believed that cross-disciplinary trainings can help to reduce many barriers to collaboration. Indeed, some studies have shown significant changes in attitude and knowledge of training participants through such educational efforts (Jones, Packard, and Nahrstedt, 2002; Mills and Yoshihama, 2002).

Building upon these recent efforts, this paper describes a statewide initiative to improve inter-agency collaboration through a series of cross-disciplinary trainings. The impact of these trainings on self-reported collaboration among a sample of child protective service workers is also examined. Given that prior research has emphasized the importance of staff attitudes and knowledge for fostering collaboration, this evaluation centers on whether these staff characteristics can be influenced by these trainings. Attention is also given to the identification of barriers and their potential role in shaping collaboration. To date, few studies have focused on the role of barriers in curtailing cooperative efforts across agencies. By providing participants with a greater understanding of the issues surrounding domestic violence and child victimization and bringing disciplines together, it is hoped that attitudes and perceptions of barriers can be positively changed; thereby, leading to greater levels of collaboration across agencies.

LITERATURE REVIEW

Cross-Disciplinary Training and Collaboration

In recent years, there have been a number of attempts to educate personnel through the use of cross-disciplinary approaches. As a result, a variety of curricula have been published addressing the delivery of such trainings to improve collaboration among child protective workers, domestic violence advocates, and other system actors (Saunders and Anderson, 2000). In 1999, the National Council of Juvenile and Family Court Judges (NCJFCJ) published a comprehensive guide titled, Effective Intervention in Domestic Violence and Child Maltreatment Cases: Guidelines for Policy and Practice, also commonly referred to as the "Greenbook." A chief principle of the "Greenbook" is that "child protective services, domestic violence programs, and juvenile courts must be committed to building internal capacity to respond effectively to families in which dual forms of maltreatment exist" (NCJFCJ, 1999, p. 25). Based on this principle, the "Greenbook" recommends that communities begin to cross-train their child welfare, domestic violence, and juvenile court system personnel.

As a result, there have been a number of recent efforts to conduct inter-agency trainings and develop curricula to better meet the challenge of enhancing collaboration in co-occurrence cases. To date, much of the curricula have focused on assessment of risk and danger, planning, and intervention methods (Mills and Yoshihama, 2006). However, others have cited the importance of incorporating information on common policies and procedures to better educate participants on the roles and responsibilities of partnering agencies and potentially identify barriers to collaboration (Appel and Appel, 2006). In theory, the goals of such cross-disciplinary trainings is to reduce barriers, overcome communication issues, create a safe working environment for all workers, and ultimately end the generational chain of violence (Button and Payne, 2009; Potito, et al., 2009).

Studies on specific curricula and training programs are rare, but a few states have initiated and evaluated the impact of trainings (Mills and Yoshihama, 2006). For example, some states have made significant progress by promoting close working relationships through cross-training (Whitney and Davis, 1999; Findlater and Kelly, 1999b). Massachusetts was the first to establish a working relationship with domestic violence and child welfare agencies when the Department of Social Services began to mandate training for all new workers (Whitney and Davis, 1999). They combined the foundations of both child welfare and domestic violence in the training. Through a pilot program they were able to create a strategy that would work with both domestic violence advocates and child welfare workers. This collaboration did not come easily for either group, but over time respect was gained and challenges were fewer. In the end, Massachusetts' efforts were successful at raising awareness and fostering working relationships between domestic violence and child welfare programs and, ultimately, resulted in the passage of critical legislation to assist in the handling of these cases.

Similar efforts have also taken place in other states. In the state of Michigan, Saunders and Anderson (2000) reported the results of a training conducted over a two day period. A published training curriculum, Domestic Violence: A National Curriculum for Child Protective Services was used to cross-train staff on the handling of co-occurrence cases (see Ganley and Schechter, 1996). Specifically, the training covered topics related to common factors associated with domestic violence and child abuse and neglect, the dynamics of domestic violence, and the identification, assessment, and practical implications of co-occurrence cases (Saunders and Anderson, 2000). Using a pre-post training survey, the evaluation sought to determine whether increasing the knowledge of child welfare workers on issues of domestic violence would influence the handling of such cases. The study reported substantive changes in child welfare workers' use of assessment, the verification of emotional abuse, and less of a tendency to hold victims responsible for stopping the violence. While the results of this evaluation were generally positive, the authors concluded that cross-training by itself may not be sufficient to remedy all of the challenges associated with educating staff on the handling of domestic violence cases.

The counties of Los Angeles and Orange in California have also sought to improve the handling of co-occurrence cases through training (Mills and Yoshihama, 2002). Using an interdisciplinary team of university faculty members and representatives of the Los Angeles County Domestic Violence Council Training Committee, the training focused on helping children's services workers "understand the complexities involved in working with families suffering from duel abuses and to develop skills in methods for assessment and intervention" (Mills and Yoshihama, 2002, p. 567). Comprised of a single day program (i.e., One-Day Program) as well as a more comprehensive single day program occurring over a period of six months for select workers and supervisors (i.e., Fellows Program), the trainings consisted of didactic teaching, role-playing exercises, as well as leadership training. Based on the responses of a pre-post training questionnaire, Mills and Yoshihama (2002) concluded that both programs were generally effective at bringing about changes in attitudes and competency among children's services workers.

Although these studies indicate that such trainings can be successful at changing workers' attitudes and knowledge, others contend that sustained efforts involving personnel training and actual policy and/or procedural change are necessary to meaningfully impact levels of inter-agency collaboration. Appel and Appel (2006) argue that there must be partnerships between agencies which address needs on an ongoing basis while, at the same time, appreciating the policy and procedures of the other agencies. As part of the Greenbook initiative, six demonstration sites were chosen to implement many of the recommendations for improving the handling of co-occurrence cases (Caliber, 2004; 2005). Each of these projects consisted of representatives from the courts, child welfare system, and domestic violence service providers and sought to improve the ways in which different agencies work together to address issues related to co-occurrence cases. These sites engaged in ongoing education and cross-training as well as made significant strides in policy and procedural development in an effort to reduce conflicts and encourage widespread community participation (Banks, Landsverk, and Wang, 2008). As a result, the evaluation team found that there was a significant decrease in the co-occurrence of domestic violence and child abuse across these six communities (ICF International, 2008). According to the project stakeholders involved, however, one of the most notable successes was the improved relationship between the domestic violence advocates and child welfare workers (Banks, Dutch, and Wang, 2008).

Barriers to Collaboration

While many efforts have been made to increase collaboration through cross-disciplinary training, others have argued that such strategies can be undermined by the presence of barriers. Because of the inherent differences between domestic violence programs and child protection services, some researchers have pointed out that training is an essential component of attempts to enhance collaboration, but is often not sufficient by itself (Appel and Appel, 2006). Acknowledging that this could be a problem, the authors of the "Greenbook" noted that such strategies would have to be implemented over a long period of time to allow for ample trust and systems of communication to be established (Janczewski, Dutch, and Wang, 2008). In large part, this may be due to the existence of system- and individual-level barriers that serve to impede cooperative efforts across agencies (Button and Payne, 2009).

Whitney and Davis (1999, p. 162) point out that "large bureaucracies, with staff turnover, inadequate funding, constant public scrutiny, and shifting political agendas" are all reasons for concern when bringing groups together. Hence, changing policy and other system factors are likely to be crucial for overcoming such obstacles. As part of a recent study of social work agencies in Virginia, for example, Button and Payne (2009) explored potential barriers to the delivery of cross-disciplinary trainings among a statewide sample of child protection supervisors. Over one-half of respondents indicated that lack of time and staff as well as travel distances posed significant barriers to the delivery of cross-disciplinary trainings. Based on such findings, the authors suggested that formal policy and guidelines are necessary to reduce barriers as well as gaps in knowledge and communication (Button and Payne, 2009).

Apart from systematic barriers, there are also individual-level barriers including staff attitudes that should be addressed through education efforts. Qualitative interviews with domestic violence workers in New York revealed concern over attitudes, which ultimately hindered collaborative efforts (Magen, Conroy, Hess, Panciera, and Simon, 2001). While it is believed that individual-level barriers may be more easily overcome compared to system-level barriers, it is critical for system impediments to be addressed in order to sustain any efforts to build collaboration. Yet, there have been few studies that have systematically examined the impact of barriers on interagency collaboration in co-occurrence cases to date.

METHODS

Project Description

With a grant from the U.S. Department of Justice, Office on Violence Against Women (OVW), a training program was developed to address the issue of the co-occurrence of domestic violence and child victimization. Supported by OVW's Rural Domestic Violence and Child Enforcement Grant Program funds, the state Coalition Against Domestic Violence (CADV) developed a comprehensive curriculum rooted in extant research on inter-disciplinary collaboration and recommendation from the National Council of Juvenile and Family Court Judges, Family Violence Department's "Greenbook" initiative. However, in addition to improving services to traditionally underserved communities through OVW's rural grant funds, one of the goals of the project was to research, educate, and train on the co-occurrence of domestic violence and child victimization. Toward that end, the CADV began its efforts to address the issue by forming a study and policy workgroup consisting of representatives from the courts, child protective services (CPS), victim advocates, and law enforcement.

Using the aforementioned resources and a series of educational forums offered by national experts, the workgroup determined cross-training among child protective services, domestic violence advocators, and others were needed to enhance collaboration and improve services to victims. Established by the CADV, the workgroup and a smaller training subcommittee developed a curriculum rooted in available research to cross-train various disciplines on each other's perspectives, roles, and responsibilities in responding to cases of family violence. The curriculum was to be delivered by a multidisciplinary training team--to a multidisciplinary audience--of domestic violence advocates, child protective service workers, law enforcement officers, and court representatives. Under a cooperative agreement between the CADV and the Department of Health and Human Resources (DHHR) training division, a series of 10 regional cross-disciplinary workshops were conducted throughout the state.

The curriculum was designed to provide participants with an understanding of the dynamics of domestic violence, the process that follows a report of child abuse and/or neglect, and the impact on families when these problems co-occur. Participants were also exposed to the guiding principles of the three main systems (i.e., child protective services, domestic violence services, and courts) as well as law enforcement. Each system was described in terms of their respective roles and responsibilities, risk assessment, and safety planning.

In addition, information on how conflict can occur via differences in the underlying principles of the three systems (and law enforcement) was presented. Participants utilized this information to develop strategies for addressing areas of contention across the systems, to identify potential barriers to collaboration, and to generate solutions. Through this process it was hoped that participant's knowledge of the roles and responsibilities as well as attitudes toward other co-occurrence victimization responders would increase; thereby, resulting in greater collaboration. Specific topics covered during the training included: the impact of batterers on adult victims, the public policy principles essential to the legal/court system, an overview of the abuse and neglect process (civil and criminal), the impact of children's exposure to batterers, assessing the risk across disciplines, planning for safety across disciplines, the roles and responsibilities of other players, bringing the players together when public policy principles conflict, and coordinating community responses. Generally, participants were expected to begin the process of identifying pathways for overcoming real and perceived barriers and, at the same time, gain an appreciation for the importance of developing inter-agency collaboration in co-occurrence cases.

Study Design and Participants

Data to evaluate the impact of the co-occurrence training curriculum were collected through the administration of a survey. A survey was given to child protective workers drawn from four of the ten regional workshops who represented various regions of the state. Employment addresses for participants were obtained from training registration lists compiled by the DHHR for the purposes of administering the post-training survey. An amended version of Dillman's total design method involving follow-up postcards and subsequent mailings of the survey were used to maximize the response rate for this sample.

The survey respondents represented two independent samples. The first sample of participants completed the survey instrument on-site prior to the start of the first training module to establish baseline knowledge levels previous to training. This sample represents the study's comparison group (i.e., a sample of CPS workers not exposed to the training information). The survey instrument was then mailed to the second post training sample six months after they had attended the training. This group represents the post training or treatment group (i.e., a group of CPS workers who were exposed to the training information). While a pre-and- post survey design would have been optimal, sampling and survey implementation issues only allowed for a comparison group design. Level of experience and demographic characteristics were also collected to better ensure that the comparison and post-training (i.e., treatment) group participants were similar.

As shown in Table 1, the two samples of CPS workers total to 146 total respondents. At the time of training roughly 380 CPS employees were working for the state meaning that the total sample represents 38% of the total population. The two groups were statistically similar on various demographic characteristics since no significant differences in the gender, race, age, or education level were found between the two groups. Participants in both the post-training group and the comparison group were predominately white females with a college degree and below 39 years of age. Likewise, over ninety percent of the comparison group and post-training participants were white and female. The average age of pre-training participants was 35 years of age, while the mean age for post-training participants was 38 years. Over eighty-five percent (85%) of the two samples had earned a college degree. A similar number of comparison and post-training participants had earned a post-graduate degree, at 6 and 5 participants respectively.

MEASURES

The survey instrument was designed to measure four main constructs: knowledge, attitudes, collaboration, and barriers--prior to and after the trainings were delivered. To measure knowledge, participants were asked to indicate the extent to which they understood the legal and/or procedural roles and responsibilities of domestic violence advocates (DV advocates), law enforcement officers, and court personnel (1) Responses were measured on a five-point Likert format with responses ranging from "not at all" to "a great deal." A mean score was then calculated for responses pertaining to each group.

The attitudes of child protective workers toward collaboration with their inter-agency partners were measured by assessing their opinions about recent collaborations. Specifically, participants were asked to indicate the degree to which they had a positive or negative view of their collaborations with each of the three groups over the past six months. Responses were measured using a five-point Likert scale ranging from "very negative" to "very positive."

For the purposes of this study, a single item was used to measure self-reported levels of collaboration. Of particular interest was the extent to which they were actually collaborating with their inter-agency partners. Respondents were asked how often they had collaborated with DV advocates, law enforcement, and court personnel over the past six months. (2) Response categories ranged from "never" to "daily," with other options including "a few times," "monthly," and "weekly." A mean score was calculated indicating the level of collaboration with each of the three groups.

Finally, child protective workers were asked about their perception of the presence or absence of barriers to collaboration. A list of potential barriers was presented to participants in an effort to ascertain which items they perceived to be more or less of a hindrance to successful collaborations between agencies. The list included a variety of system and individual-level barriers that child protective workers may have encountered when seeking to collaborate with other agency representatives. System-level barriers pertained to factors related to the policy and/or procedures and daily operational aspects of bureaucratic agencies. These barriers included items such as "high turnover rates for workers," "agency policies and/or procedures," "confidentiality restrictions or requirements," and so forth. Individual-level barriers consisted of such items as "lack of interpersonal relationships," "differences in ideological values," and "lack of confidence in counterpart's knowledge." Respondents were asked to rate the extent to which they felt each of the items represented barriers to collaborative efforts in cases involving the co-occurrence of domestic violence and child victimization. Each item was measured on a five-point scale ranging from "not at all" to "a great deal."

ANALYSIS

Results of independent-samples t-tests were used to determine if the training was effective in increasing knowledge, attitudes, and collaboration post-training. The t statistic was also used to explore the post-training system and individual-level barriers as well as whether the training influenced child protective workers' perceptions of such barriers. Pearson correlation coefficients were calculated to examine the relationship of collaboration among domestic violence advocates, law enforcement, and courts.

RESULTS

The presentation of the results begins with an examination of post-training changes in participants' knowledge levels, attitudes toward other agency representatives, and levels of collaboration. These findings are followed by an assessment of the relationship between knowledge, attitudes, and collaboration for the post-training sample of participants. Lastly, post-training perceptions of system and individual-level variables are explored.

Changes in Participants' Knowledge, Attitudes, and Level of Inter-Agency Collaboration

Table 2 displays the mean changes in knowledge, attitude, and self-reported collaboration of the comparison and post-training (or treatment) groups using a series of independent-samples t-tests. The findings indicate that the training did not result in statistically significant changes in the mean levels of these measures. However, there is evidence that the training resulted in some improvements and that these changes varied across domestic violence advocates, law enforcement personnel, and court representatives. The comparison group's baseline measures show that CPS workers were less knowledgeable of the legal roles and/or procedural responsibilities (mean = 3.13) and reported less collaboration with domestic violence advocates (mean = 2.03) compared to other agency counterparts. Surprisingly at post-training, an examination of the mean scores showed that there were modest decreases in CPS worker's rating of collaboration with all other agencies and frequency of collaboration with domestic violence advocates; however, none of these differences were statistically significant. Nevertheless, mean scores for knowledge, attitudes, and collaboration remained lower for domestic violence advocates compared to law enforcement personnel and court staff. Clearly, CPS workers were more knowledgeable, held more favorable attitudes, and collaborated more often with law enforcement and court personnel prior to and after the training.

Relationship between Knowledge, Attitudes, and Collaboration at Post-Training

Table 3 illustrates that the knowledge and attitudes of CPS workers toward other co-occurrence partners (domestic violence advocates, law enforcement, and court personnel) differed at baseline and post-training. As expected, knowledge of the legal roles and responsibilities of other co-occurrence partners and attitudes based on prior collaborations were shown to be more favorable in the post-training sample in most cases as there were statistically significant correlations with CPS workers' self-reported levels of collaboration. Both composite measures between the knowledge and attitudes of CPS workers and levels of collaboration were statistically significant in the comparison and post-training sample. At the same time, there were significant differences in attitudes toward collaboration found in the post-training sample which were not observed in the comparison sample. This provides evidence that post training that the frequency of collaboration strengthened the knowledge and positive attitudes toward interagency collaboration with the exception of domestic violence advocates. While Table 2 showed no overall increase in collaboration post training, this finding suggests that those who have increased the collaboration learn more about other agencies and feel better about their collaboration with the exception of domestic violence advocates.

The comparison group reported significant baseline correlations between frequency of collaboration and knowledge for all co-occurrence partners and the composite measure. The relationship between CPS workers' knowledge of the legal roles and responsibilities was highest for the knowledge composite measure (r=.338, p=.004). However, correlations were higher in the post training sample for every knowledge of roles and responsibilities measure except domestic violence advocates which went down from (r = .309, p = .008) in the comparison group to (r = .229, p = .013) in the post training sample. In the post training sample the correlation between frequency of collaboration and knowledge of court personnel was the strongest (r = .439, p = .000) up from (r = .268, p = .023) in the comparison group. There was also an increase in the correlation between knowledge and frequency of collaboration post training from (r = .338, p = .004) in the comparison group to (r = .428, p = .000).

Frequency of collaboration and attitudes toward collaboration were only found to be correlated with domestic violence advocates (r = .316, p = .007) and the composite measure (r = .268, p = .023) in the comparison group. There were no significant correlations between collaboration and attitudes toward collaboration for law enforcement and court personnel. However, in the frequency of collaboration was found to be strongly related to positive attitudes toward working with law enforcement (r =.418, p =.000) and court personnel (r = .351, p =.003) in the post training sample. The overall correlation between frequency of collaboration and the composite attitude measure (r =.420, p =.000) post training was the strongest finding for the attitude measures and higher than the attitude composite in the comparison group (r = .268, p =.023) These findings imply that CPS workers surveyed after the training had a better relationship with law enforcement and court personnel than a similar sample of CPS workers did before the interagency training

class.

Changes in Participants' Perception of System and Individual-level Barriers to Collaboration

Table 4 displays the changes in CPS workers' perception of system and individual-level barriers to collaboration. The results indicate that CPS workers perceived multiple barriers to collaboration at both levels. For the most part, a greater proportion of CPS workers viewed barriers to be related to system-level factors. Such system-level factors as high turnover rates, time constraints, and too few staff were perceived to be important barriers prior to and after the training. Over sixty percent of CPS workers in both groups reported these to be important barriers to collaboration. On the contrary, forty percent or fewer of respondents viewed individual-level barriers to be important in curtailing collaboration. The results further indicate that the training may have been successful at reducing the importance of nearly all of the barriers identified by CPS workers. For instance, all of the system-level barriers declined in importance or stayed the same at post-training, except for high turnover rates which stayed the same. In terms of system factors, there were substantial reductions in the proportion of respondents indicating that "too few staff," "differences in agency mandate and "lack of contact between agencies" presented significant barriers to collaboration. Seventy-three percent (72.9%) of respondents reported that "too few staff" represented a major barrier to collaboration prior to the training, compared to 64.3% at post-training group. In the same regard, over forty percent of respondents indicated that "differences in agency mandates" were a significant impediment to collaboration prior to the training (44.3%). This is almost thirteen percent greater than the proportion rating differences in agency mandates as a barrier to collaboration in the post training sample (31.4%).

The training also appeared to influence CPS workers' perceptions of many individual-level barriers to collaboration. Similar to the changes in many of the structural barriers, all individual-level barriers decreased in their importance among this sample of CPS workers, with the exception of "failed collaborations in the past. The proportion of respondents reporting that "lack of interpersonal relationships" was an important barrier was reduced by more than half at post-training. Thirty-three percent (32.9%) of CPS workers in the comparison group indicated that a "lack of interpersonal relationships" was a barrier to inter-agency collaboration, compared to only 14.5% of respondents at post-training group. The percentage of CPS workers that reported a "lack of confidence in counterparts' knowledge" also declined subsequent to the cross-disciplinary training (i.e., 34.3% to 17.1%).

Table 4 further illustrates that the rank-order of importance for many of these barriers to collaboration shifted as a result of the training.

While the top four system-level barriers at pre-training held the same ranking at post-training (i.e., high turnover rates, time constraints, too few staff, and confidentially restrictions), others became more or less important after the training. For instance, rank-order of both "agency policies and/or procedures" and "lack of contact between agencies" shifted in order of importance. As "lack of contact between agencies" became less important as a barrier, "agency policies and/or procedures" became more important after the training. Similar shifts in rank-order of importance also occurred among the individual-level barriers. Somewhat surprisingly, a "lack of interpersonal relationships" and a "lack of confidence in counterparts' knowledge"--both of which were principle targets of the co-occurrence training--declined considerably in rank order of importance among the post-training respondents. As a result, these two factors became the least important obstacles to collaboration as reported by the post-training sample of CPS workers.

Table 5 displays the results of independent-samples t-test for the comparison of perceived barriers prior to and after the cross-disciplinary training. Consistent with the findings presented in Table 4, there were significant reductions in the perception of some barriers among CPS workers. An examination of the mean scores showed statistically significant declines for "too few staff" (t = 3.011; p = .003) and "lack of contact between agencies" (t = 2.617; p = .010) as perceived barriers to collaboration. Likewise, the perception of "lack of interpersonal relationships" as an obstacle to collaborative efforts was lower in the post-training sample (t = 3.720; p = .000). The "lack of confidence in counterpart knowledge" was significantly lower in the post-training group (t=2.820; p=.006). The training did not appear to be successful at significantly reducing CPS workers' perceptions of other barriers. In particular, mean scores for many system and individual-level barriers relating to staffing (high turnover rates, confidentially restrictions, and accessibility of counterparts) and/or policy (differences in agency mandates, agency policies/procedures, time constraints, and differences in ideological values) did not significantly decline at post-training.

DISCUSSION AND CONCLUSION

One decade ago, the National Council of Juvenile and Family Court Judges (1999) initiated a project, often referred to as the "Greenbook" initiative, which was designed to establish guidelines for the development of practice and policy in cases where domestic violence and child victimization overlap. One of the basic tenants of this initiative called for child protective services providers, domestic violence programs, and juvenile courts to build their internal capacity to respond effectively to families where dual forms of maltreatment exist. Rooted in this basic tenant was the recommendation that communities should cross-train their child welfare, domestic violence, and juvenile court system personnel in an effort to enhance inter-agency collaborations and improve the handling of co-occurrence cases (Janczewski, Dutch, and Wang, 2008). The Greenbook authors noted that "cross-communication and training are the foundations on which successful collaborations can be built" (NCJFCJ, 1999, p.39).

Given the emphasis of improving inter-agency collaboration through cross-training, this study sought to examine the impact of such trainings on CPS workers' knowledge and perceptions of other co-occurrence victimization responders. Previous studies on co-occurrence and collaboration have centered on the relationship between child protective workers and domestic violence advocates (e.g., Button and Payne, 2009; Whitney and Davis, 1999; Saunders and Anderson, 2000; Mills and Yoshihama, 2002; Findlater and Kelly, 1999a; Nuszkowski, et. al., 2007). However, this study was able to compare CPS workers' knowledge of--and attitudes toward--domestic violence advocates, law enforcement, and court personnel. It was hoped that by changing CPS workers' knowledge, attitudes, and opinions of their counterparts in other agencies (i.e., domestic violence advocates, law enforcement officers, and court personnel) that inter-agency collaboration would increase.

While we did not find significant differences in CPS workers' knowledge, attitudes and frequency of collaboration after the training overall. Perhaps this training did not have the anticipated impact on knowledge, attitudes, and levels of self-reported collaboration because of the "ceiling effect." According to Saunders and Anderson (2000), the ceiling effect can reduce the likelihood of finding statistically significant results when training participants score high on such measures as knowledge and attitudes prior to the training; thereby, leaving little room for improvement. Mean scores on the knowledge and attitudes of CPS workers prior to the training suggested that they generally had a good understanding of the legal roles and/or procedural responsibilities of other system actors. Moreover, most of the training participants held somewhat favorable attitudes toward their agency counterparts as indicated by mean scores, particularly for law enforcement and court personnel. Secondly, it is important to note that this study had a six-month follow-up period for the collection of post-training surveys which is considerably longer than most previous studies that have examined the impact of similar cross-disciplinary trainings. Prior evaluations of cross-disciplinary trainings on co-occurrence contain little or no follow-up period, but instead gathered post-training data immediately or shortly after the workshops took place (e.g., Saunders and Anderson, 2000; Mills and Yoshihama, 2002).

While we did not have any significant findings in the overall measures post training our results indicated that as workers became more educated on the legal and/or procedural roles and responsibilities of agency counterparts, their level of self-reported level of collaboration also increased. The only exception was with domestic violence advocates where frequency of collaboration was related to lower levels of knowledge and positive attitudes post training. Of particular interest is that collaboration was only found to be related to positive attitudes toward law enforcement and court personnel in the post-training sample but was not related in the comparison group. These findings relative to law enforcement and court personnel are consistent with other studies that have highlighted the importance of such factors for increasing collaboration across agencies (Mills and Yoshihama, 2002; Magen, et. al., 2001; Jones, et. al., 2002). This suggests that while the overall frequency of collaboration may stay the same for most CPS workers the knowledge gained and positive attitudes toward collaboration increases with collaboration post training.

This current study also produced valuable information on how cross-disciplinary trainings may influence the perception of barriers to collaboration on the part of CPS workers. In a national survey of multidisciplinary team leaders, Kolbo and Strong (1997) found initial and ongoing training to be the most frequently cited strategies for overcoming barriers toward collaboration. Yet, few studies have sought to assess whether such trainings can be successful at reducing worker perceptions of such barriers. Generally, we discovered a host of structural and individual-level factors that may serve to inhibit collaboration in co-occurrence cases. In particular, a large proportion of respondents identified structural-level factors (that is, factors related to the policy and/or procedures and daily operational aspects of bureaucratic agencies) to be impacting inter-agency collaboration "quite a bit" or "a great deal." Over two-thirds of CPS workers identified such structural barriers as high turnover rates, time constraints, and too few staff as being substantial barriers prior to and after the training. It is not entirely clear whether these barriers can be adequately addressed solely through cross-disciplinary trainings.

Nonetheless, our analysis found that cross-disciplinary trainings can change the way CPS workers view the presence of most barriers. In examining comparison group and post-training responses, we found that virtually all system and individual-level barriers declined in importance after receiving the cross-disciplinary training. Moreover, the results highlighted changes in the rank order of most barriers prior to and after the training. Consistent with the goals and objectives of the training, there were declines in the perceived importance of barriers related to inter-agency contacts and relationships at post-training. As indicated by the results of the independent samples t-tests, significant reductions in CPS worker perceptions of "lack of contact between agencies" and "lack of interpersonal relationships" may have occurred as a result of the training. In addition, respondents were significantly less likely to see "too few staff" as an important factor in inhibiting collaborative efforts across agencies. These results imply that such trainings can have an impact on the perceptions of such barriers; however, it is not yet known how these changes translate into actual levels of collaboration. Future research should systematically explore the role of these barriers and their impact on the handling of co-occurrence cases and work to identify optimal strategies for dealing with such obstacles to collaboration.

Although the impact of the training was limited in its capacity to produce the desired changes of increased collaboration, this research helped to broaden our understanding of the relationships between CPS workers and other system actors and barriers to collaboration. We found that CPS workers reported being more knowledgeable of the legal and/or procedural roles of law enforcement and court personnel compared to domestic violence advocates and that knowledge and attitudes improved with frequency of collaboration post training. These findings imply that more work is needed to build successful linkages and partnerships, especially between domestic violence and child welfare agencies.

REFERENCES

Appel, J.K. & Appel, D.K. (2006). Child maltreatment and domestic violence: Human services issues. Journal of Health and Human Services Administration, 29(2), 228-244.

Appel, A.E. & Holden, G.W. (1998). The Co-occurrence of spouse and physical child abuse: A review and appraisal. Journal of Family Psychology, 12 (4), 578-599.

Banks, D., Dutch, N., & Wang, K. (2008). Collaborative efforts to improve system response to families who are experiencing child maltreatment and domestic violence. Journal of Interpersonal Violence, 23 (7), 876-902.

Banks, D., Hazen, A. L., Jeffrey, C. H., Wang, K., & Griffith, J. D. (2008). Collaboration between child welfare agencies and domestic violence service providers: Relationship with child welfare policies and practices for addressing domestic violence. Children and Services Review, 31, 497-505.

Banks, D, Landsverk, J., & Wang, K. (2008). Changing policy and practice in the child welfare system through collaborative efforts to identify and respond effectively to family violence. Journal of Interpersonal Violence, 23 (7), 903-932.

Button, D.M. & Payne, B.K. (2009). Training child protective services workers about domestic violence: Needs, strategies, and barriers. Children and Youth Services review, 31 (3), 364-369.

Caliber Associates. (2004). The Greenbook demonstration initiative process evaluation report: Phase I (January 2001 - June 2002).

Caliber Associates. (2005). The Greenbook demonstration initiative interim evaluation report.

Edleson, J.L. (1999). The overlap between child maltreatment and woman battering. Violence Against Women, 5 (2), 134-154.

Findlater, J.E. & Kelly, S. (1999a). Child protective services and domestic violence. The Future of Children, 9 (3), 84-96.

Findlater, J.E. & Kelly, S. (1999b). Reframing child safety in Michigan: Building collaboration among domestic violence, family preservation, and child protection services. Child Maltreatment, 4 (2), 167174.

Ganley, A.L. & Schechter, S. (1996). Domestic violence: A national curriculum for children's protective services. San Fancisco, CA: Family Violence Prevention Fund.

ICF International (208). The Greenbook Initiative: Final evaluation report. Fairfax, VA: Author.

Janczewski, C., Dutch, N., & Wang, K. (2008). Crafting the Greenbook: Framers reflect on the vision, process, and lessons, learned. Journal of Interpersonal Violence, 23 (7), 981-1006.

Jones, L., Packard, T., and Nahrstedt, K. (2002). Evaluation of a training curriculum for inter-agency collaboration. Journal of Community Practice, 10 (3), 23-40.

Kolbo, J.R. & Strong, E. (1997). Multidisciplinary team approaches to the investigation and resolution of child abuse and neglect: A national study. Child Maltreatment, 2 (1), 61-72.

Magen, R.H., Conroy, K, Hess, P.M., Panciera, A., & Simon, B.L. (2001). Identifying domestic violence in child abuse and neglect investigations. Journal of Interpersonal Violence, 16 (6), 580-601.

Mills, L.G. & Yoshihama, M. (2002). Training children's services workers in domestic violence assessment and intervention: Research findings and implications for practice. Children and Youth Services Review, 24 (8), 561-581.

Moles, K. (2008). Bridging the divide between child welfare and domestic violence services: Deconstructing the change process. Children and Youth Services Review, 30, 674-688.

National Council of Juvenile and Family Court Judges (NCJFCJ). (1999). Effective intervention in domestic violence & child maltreatment cases: Guidelines for policy and practice. Reno, Nevada: Author.

Nuszkowski, M.A., Coben, J.H., Kelleher, K.J., Goldcamp, J.C., Hazen, A.L., & Connelly, C.D. (2007). Training, co-training, and cross-training of domestic violence and child welfare agencies. Families in Society: The Journal of Contemporary Social Services, 88 (1), 35-41.

Potito, C., Day, A., Carson, E., & O'Leary, P. (2009). Domestic violence and child protection: Partnerships and collaboration. Australian Social Work, 62 (3), 369-387.

Saunders, D.G. & Anderson, D. (2000). Evaluation of a domestic violence training for child protection workers & supervisors: Initial results. Children and Youth Services Review, 22 (5), 373-395.

Spears, L. (2000). Building bridges between domestic violence organizations and child protective services. Harrisburg, PA: National Resource Center on Domestic Violence.

Whitney, P. & Davis, L. (1999). Child abuse and domestic violence in Massachusetts: Can practice be integrated in a public child welfare setting? Child Maltreatment, 4 (2), 158-166.

(1) Responses relating to defense attorneys, prosecutors, and judges were grouped together to form the court personnel category.

(2) It is important to note that "over the past six months" was used to coincide with the follow-up period applied in this evaluation.

STEPHEN M. HAAS

SIMON BAUER-LEFFLER

ERICA TURLEY

West Virginia Office of Research and Strategic Planning
Table 1
Demographic characteristics of comparison and post-training samples

                  Control Sample       Post-training
                      (N=75)              Sample
                                          (N=71)

                          [%.bar]             [%.bar]
                [n.bar]   (a)       [n.bar]   (a)

Gender
Male            17        23.6%     12        16.9%
Female          55        76.4%     59        83.1%
Total           72        100.0%    71        100.0%

Race
White           66        93.0%     67        97.1%
Nonwhite        5         7.0%      2         2.9%
Total           71        100.0%    69        100.0%

Education
High Sch.       1         1.4%      3         4.3%
Associate       1         1.4%      2         2.9%
Bachelor        64        88.9%     60        85.7%
Graduate        6         8.3%      5         7.1%
Total           72        100.0%    70        100.0%

                 Mean   SD      Mean    SD

Total Years      1.60   .493    1.59    .495

of Experience

Respondent Age   34.80  8.743   38.12   12.10

(a) may not total 100% due to rounding.

Table 2
Independent-samples t-test of changes in Knowledge, Attitudes,
and Self-reported Frequency of Collaboration

                                   Comparison (N=75)

Measures             [n.bar]   [mean.bar]   [SD.bar]   [n.bar]

Knowledge of legal
roles and/or
responsibilities

Domestic violence      70         3.13        0.93       73
advocates

Law enforcement        71         3.56        1.05       73
officers

Court personnel        71         3.53        1.07       73
Attitude toward
collaboration

Domestic violence      68         3.79        0.84       72
advocates

Law enforcement        69         4.16        0.72       71
officers

Court personnel        71         3.99        0.68       72
Frequency of
collaboration

Domestic violence      68         2.03        0.93       71
advocates

Law enforcement        69         2.72       1.056       74
officers

Court personnel        70         2.92       1.001       74

                               Post-training (N=71)

                                                       [bar.p
Measures             [mean.bar]   [SD.bar]   [t.bar]   value]

Knowledge of legal
roles and/or
responsibilities

Domestic violence       2.84        1.01      -0.07     .944
advocates

Law enforcement         3.58        0.97      -0.11     .915
officers

Court personnel         3.54        0.91      1.06      .292
Attitude toward
collaboration

Domestic violence       3.64        0.89      1.06      .292
advocates

Law enforcement         3.93        0.93      1.63      .105
officers

Court personnel         3.86        0.75      1.06      .291
Frequency of
collaboration

Domestic violence       1.93       0..95      0.63      .532
advocates

Law enforcement         2.96       1.176      -1.25     .212
officers

Court personnel         3.12       1.110      -1.10     .275

Table 3

Comparison and post-training correlations between knowledge,
attitudes, and CPS worker Collaboration

CPS Workers Frequency of Collaboration

                          Comparison (N=75)

                                         [p value
Measures             [n.bar]   [r.bar]    .bar]

Knowledge of legal
roles and/or
responsibilities

Domestic violence      72      .309 **     .008
advocates

Law enforcement        72      .274 *      .020
officers

Court personnel        72      .268 *      .023

Composite              72      .338 **     .004
Attitude toward
collaboration

Domestic violence      72      .316 **     .007
advocates

Law enforcement        71       .130       .281
officers

Court personnel        72       .138       .247

Composite              72      .268 *      .023

                     Post-training (N=71)

                                         [p value
Measures             [n.bar]   [r.bar]    .bar]

Knowledge of legal
roles and/or
responsibilities

Domestic violence      69      .299 *      .013
advocates

Law enforcement        70      .346 *      .003
officers

Court personnel        70      .439 **     .000

Composite              70      .428 **     .000
Attitude toward
collaboration

Domestic violence      67      .249 *      .042
advocates

Law enforcement        68      .418 **     .000
officers

Court personnel        70      .351 **     .003

Composite              70      .420 **     .000

Table 4

Rank-order frequency and percentage of CPS workers'
perception of system and individual level barriers

Barriers (Comparison) (c)                           f    % (b)

System-level

High turn-over rates for workers (1)                56   80.0
Time constraints (2)                                52   74.3
Too few staff (3)                                   51   72.9
Confidentiality restrictions or requirements (4)    32   45.7
Differences in agency mandates (5)                  31   44.3
Lack of contact between agencies (6)                29   41.4
Agency policies/procedures (7)                      28   40.0

Individual-level

Different priorities in the handling of cases (1)   28   40.0
Accessibility of counterparts (2)                   27   38.6
Lack of confidence in counterpart knowledge (3)     24   34.3
Differences in ideological values (4)               23   32.9
Lack of interpersonal relationships (4)             23   32.9
Inability to agree on actions to be taken (5)       22   31.4
Failed collaborations in the past (6)               20   29.0
Barriers (Post-Training)                            f    %

System-level

High turn-over rates for workers (1)                56   80.0
Time constraints (2)                                46   66.7
Too few staff (3)                                   45   64.3
Confidentiality restrictions or requirements (4)    32   37.1
Differences in agency mandates (6)                  22   31.4
Lack of contact between agencies (7)                29   27.1
Agency policies/procedures (5)                      23   33.3

Individual-level

Different priorities in the handling of cases (1)   23   32.9
Accessibility of counterparts (1)                   23   32.9
Lack of confidence in counterpart knowledge (5)     12   17.1
Differences in ideological values (4)               14   20.3
Lack of interpersonal relationships (6)             10   14.5
Inability to agree on actions to be taken (2)       21   30.0
Failed collaborations in the past (3)               19   27.5

(a) Rank order indicated in parenthesis.
(b) Frequencies and percentages pertain to responses of "quite a
bit" and "a great deal" as a perceived barrier.
(c) For all comparison items (n = 70), failed collaborations in the
past (n = 69). (d) For all post-training items (n = 70), time
constraints, agency policies, failed collaborations, lack of
interpersonal relationships and differences in ideological values
(n = 69).

Table 5
Independent samples t-test of perceived system and individual
level barriers

                                 Comparison

Barriers               [n.bar]   [mean.bar]   [SD.bar]

System-level
High turn-over         70        4.19         .937
rates for workers

Time constraints       70        4.19         .856

Too few staff          70        4.23         .887

Confidentially         70        3.44         1.002
restrictions or
requirements

Differences in         70        3.40         .954
agency mandates

Lack of contact        70        3.47         .863
between agencies

Agency                 70        3.36         .885
policies/procedures
Individual-level

Accessibility of       70        3.39         .728
counterparts

Different priorities   70        3.33         .775
in handling of
cases

Lack of                70        3.24         .842
interpersonal
relationships

Lack of confidence     70        3.23         .920
in counterpart
knowledge

Differences in         70        3.23         .871
ideological values

Failed                 69        3.10         .877
collaborations in
the past

Inability to agree     70        3.13         .815
on actions to be
taken

                               Post-training

Barriers               [n.bar]   [mean.bar]   [SD.bar]

System-level
High turn-over         70        4.20         .844
rates for workers

Time constraints       69        3.93         .960

Too few staff          70        3.74         1.017

Confidentially         70        3.40         .907
restrictions or
requirements

Differences in         70        3.19         .839
agency mandates

Lack of contact        70        3.09         .880
between agencies

Agency                 69        3.16         .949
policies/procedures
Individual-level

Accessibility of       70        3.21         .797
counterparts

Different priorities   70        3.30         .922
in handling of
cases

Lack of                69        2.71         .842
interpersonal
relationships

Lack of confidence     70        2.80         .878
in counterpart
knowledge

Differences in         69        2.96         .915
ideological values

Failed                 69        3.06         .968
collaborations in
the past

Inability to agree     70        3.04         .970
on actions to be
taken

                                 [p value
Barriers               [t.bar]   .bar]

System-level
High turn-over         -.095     .925
rates for workers

Time constraints       1.674     .096

Too few staff          3.011     .003 **

Confidentially         .265      .791
restrictions or
requirements

Differences in         1.411     .160
agency mandates

Lack of contact        2.617     .010 *
between agencies

Agency                 1.270     .206
policies/procedures
Individual-level

Accessibility of       1.329     .186
counterparts

Different priorities   .198      .843
in handling of
cases

Lack of                3.730     .000 ***
interpersonal
relationships

Lack of confidence     2.820     .006 **
in counterpart
knowledge

Differences in         1.796     .075
ideological values

Failed                 .276      .783
collaborations in
the past

Inability to agree     .566      .572
on actions to be
taken
Gale Copyright: Copyright 2011 Gale, Cengage Learning. All rights reserved.