Child Welfare Research and Training

Social Work Education

Of course, social work education should always be seeking ways to improve generally and particularly in educating future child welfare practitioners, administrators, and researchers. Nonetheless, contrary to Stoesz’ suggestion, the publication rates of social work faculty, deans, and CSWE board members are not the best indicators of whether the profession is suited to educate and prepare quality child welfare workers. Studies that compare child welfare workers holding a social work degree with workers holding other degrees constitute better evidence. Such studies present a mixed picture. Some studies find that child welfare workers holding a social work degree outperform workers with other degrees in outcomes such as post-training child welfare knowledge or skills (Bagdasaryan, 2012; Franke, Bagdasaryan, & Furman, 2009). Moreover, convincing studies have found that compared to workers with other degrees, social workers are more likely to investigate and substantiate maltreatment, more likely to place children with relatives and adoptive homes, less likely to place children in residential settings (Barbee et al., 2009), and more likely to achieve family reunification (Ryan, Garnier, Zyphur, & Zhai, 2006).

Studies comparing social work to other disciplines are not uniformly favorable to social work, however. One study found that social workers scored higher than child welfare workers with other degrees in measures of knowledge and skill at graduation, but after three years of practice, social workers looked similar to nonsocial workers in such measures (Scannapieco, Hegar, & Connell-Carrick, 2012). Whereas some researchers found that child welfare workers with social work degrees had better performance evaluations than workers with other degrees (Rosenthal & Watters, 2006), one study found that child welfare performance review evaluations were similar for supervisors with social work degrees and other degrees (Perry, 2006). It will be important to continue monitoring such comparisons with strong study designs and methods. In addition, going forward, it will be important to use strong study designs to assess outcomes attained from child welfare workers exposed to various training curricula and methods as well as alternative degree programs.

Stoesz specifically argues for better research training of social work students and cites CSWE’s failure to require a research thesis as evidence of poor research education in social work. We also support strong research training in social work, but rather than a thesis requirement, we think research education will be strengthened more effectively through curricula that engage students in hands-on projects requiring application of rigorous methods. Thus, we are pleased to see the NCWWI’s tangible contributions to enhancing social work research training. One of the efforts funded by the NCWWI, for example, is a university–agency partnership involving the University of California at Berkeley School of Social Welfare. One component of the Cal-Child Welfare Leadership Training (Cal-CWLT) partnership involves an innovative, rigorous curriculum in which Master of Social Work (MSW) students with child welfare career interests are trained to analyze and interpret child welfare administrative data and to identify ways to use data to inform child welfare practice. With the opportunity to work with actual child welfare data, students develop skills to use data to inform practice before they enter the workplace (Lery, Putnam-Hornstein, Wiegmann, & King, 2015; Shaw, Lee, & Wulczyn, 2012).

Stoesz also criticizes social work for embracing postmodern philosophy. Yet one professor’s 2013 award for “best conceptual article” from a social work journal notwithstanding, Stoesz, impression that the profession has embraced postmodernism is not supported by evidence. We are not aware of any analyses of, for example, articles published in leading social work journals, presentations at conferences of membership organizations, such as the Society for Social Work and Research, or research courses taught in social work programs that demonstrate an embrace of postmodern philosophy. In any event, there is surely no evidence that the NCWWI adopts a postmodern philosophy, and the critical question is whether the NCWWI supports rigorous research.

Research to Improve Child Welfare

Stoesz makes a case for funding more RCTs of child welfare interventions and expresses regret that the NCWWI funds mixed methods designs, which Stoesz regards as “less than optimal.” We also support funding more RCTs, and we particularly see a need for RCTs at the organization, county, or even state level to assess systems-change efforts. We disagree, however, that more RCTs alone will address current research needs to evaluate social work child welfare training and improve child welfare services. Further, RCTs are not incompatible with mixed methods designs but can go hand in hand. In fact, leading researchers argue for adding qualitative or other methods to RCTs as a means to help explain how and why interventions attain (or fail to attain) a desired outcome. Some argue that mixed methods designs that incorporate an RCT as one part of a study can help to inform intervention improvement and identify factors that are relevant to implementing an intervention in various practice settings (Hesse-Biber, 2012; Nelson, Macnaughton, & Goering, 2015; White, 2013). Interventions demonstrated to work in RCTs sometimes fail to demonstrate the same outcomes in practice settings. Using other methods in conjunction with RCTs can provide additional information about implementation and context that child welfare agencies can use in implementation efforts.

More importantly, for several reasons, focusing solely on RCTs is unlikely to address all research needs for improving child welfare. The idea that interventions demonstrated to work in RCTs or any other research-supported interventions (RSIs) can be simply adopted in complex practice settings, such as child welfare, is no longer widely held. Cutting-edge researchers now focus on effectively integrating RSIs within an evidence-based practice approach (Aarons, Hurlburt, & Horwitz, 2011; Proctor & Rosen, 2008), implementing RSIs in challenging settings by first preparing interpersonal networks and transforming the organizational context (Fixsen, Blasé, Naoom, & Wallace, 2009; Glisson, Green, & Williams, 2012; Palinkas et al., 2009; Palinkas et al., 2011), and building evidence through continuous use of data in practice (Testa et al., 2014; Wulczyn, Alpert, Orlebeke, & Haight, 2014). As it happens, complex approaches to the use of evidence to transform organizations and improve practice are precisely the research approaches supported by the NCWWI. Such approaches are not “sloppy and substandard” but reflect the most current thinking on how to effectively infuse evidence in challenging service systems.

As one example, in addition to the applied child welfare research training for MSW students, the NCWWI-funded Cal-CWLT runs a training program for child welfare supervisors and middle managers. Reflecting an assessment of the challenges associated with implementing RSIs in child welfare, Cal-CWLT prepares child welfare supervisors and middle managers to use data to improve practice via continuous quality improvement (CQI) (Lery, Wiegmann, & Berrick, 2015). Cal-CWLT uses a CQI model developed by researchers at the Center for State Child Welfare Data at Chapin Hall (an entity that Stoesz seems to admire). The CQI process applies steps of the scientific method to decision making in applied child welfare settings. The process also addresses the role of organizational conditions in promoting evidence-based practice. It helps practitioners to understand data, to see its relevance to practice, and to develop skills to use data to inform practice.

NCWWI as a Cartel

Stoesz criticizes the NCWWI for steering funds only to social work and keeping out other disciplines. Yet one of the five primary university partners funded by and leading the NCWWI is the University of Southern Maine’s Cutler School for Health and Social Policy, which is part of the Muskie School of Public Service, not the university’s School of Social Work. Cutler School staff members run the NCWWI’s Leadership Academy for Supervisors. Hence, the NCWWI has not excluded other disciplines from the opportunity to participate in efforts to improve child welfare.

For several good reasons, however, the institute’s funding for “traineeships” supports students pursuing social work education. Social work programs have years of commitment to child welfare practice and infrastructures to train students in child welfare internships. Top child welfare journals, while interdisciplinary, are nevertheless dominated by articles from social work researchers. Perhaps someday other disciplines will embrace the preparation of child welfare practitioners and develop partnerships with child welfare agencies as social work has, but currently no other disciplines are comparably prepared to fill the important child welfare education, training, and research needs.

Future Directions

Recognizing the need to “professionalize” the child welfare workforce, the federal government, via the Children’s Bureau, has invested considerably in partnerships between state child welfare agencies and social work programs. As David Stoesz describes, Title IV-E-funded partnerships between state child welfare agencies and universities spread through the 1980s and 1990s, resulting in US$200 million distributed to 80 social work programs by 2010. With about US$20 million to distribute among its partners, NCWWI’s portion of child welfare training is small by comparison. Given the relatively small portion of child welfare education and training funded by NCWWI, the relative size and quality of the child welfare research stemming from NCWWI-funded programs are impressive.

Partly due to NCWWI’s leadership, child welfare agencies and social work programs are making good progress in instituting greater rigor and relevance in research preparation, and in bringing evidence to child welfare practice, supervision, and administration. Nevertheless, we still need more studies with strong designs, sample sizes with adequate power, and rigorous, appropriate methods to assess the effects of child welfare training on family outcomes for families (Hartinger-Saunders & Lyons, 2013; Smith, 2002). Child welfare services and outcomes remain disappointing, and many questions remain about the critical mechanisms and intervention points for improvement. Do certain training curricula or training methods produce better outcomes for families? Are child welfare workers with certain personal characteristics, ideologies, or passions more effective? Can organizational conditions be modified to enhance performance at all levels and improve family outcomes? As we go forward, the very best way to respond to David Stoesz’ and other critics’ concerns about quality services and the use of evidence in child welfare practice will be to provide study results that convincingly demonstrate how to attain superior outcomes for children, families, and communities, and to further demonstrate that federal child welfare training funds are directed to support and implement empirically-demonstrated best practices.

For now, rather than condemn an entity that is helping universities and agency partners to make progress in these important directions, we should acknowledge and congratulate the NCWWI. Contrary to functioning as a “cartel” seeking to monopolize and profit from its critical role in research to improve child welfare training and practice, the NCWWI is a model in working collaboratively with agencies, the federal government, and universities to develop and test multilevel interventions to improve child welfare, effectively prepare students for child welfare practice, and put the best knowledge and methods we have to work.

David Stoesz ultimately calls for “a vision of children’s services congruent with the 21st century” (p. 482, this issue). He argues that “essential to such a vision is the role research plays in developing interventions that produce outcomes which are positive as well as cost effective” (p. 482, this issue). He concludes that “federal funds should be awarded to those institutions best able to train those workers committed to serving the most vulnerable of Americans, maltreated children” (p. 482, this issue). We could not agree more with these final observations. Based on its record to date, it appears as if the leaders and partners of the NCWWI strongly agree as well.