Knowledge That Changes Social Work Practice: An Exploration of its Sources and Content

There is continuing interest in the relationship between knowledge and practice in social work. Overly narrow conceptualizations of the EBP model deepened the gap between practice knowledge and formal research evidence in the profession. While much has been written about the dissemination and adaptation of research findings to practice, much less is known about the actual sources of knowledge social workers draw on in their practice. This paper reports findings from an exploratory survey about the sources and content of knowledge that changed professional practice among social work field instructors (n=250) in St. Louis. An analysis of open-ended responses revealed that coworkers and continuing education programs are the most important sources for knowledge and information that influence practice. While academic journals are perceived by practitioners to be relatively unimportant sources for such knowledge, research findings on the background and effectiveness of interventions, make up the primary content that appears to affect social work practice. The findings suggest that formal research knowledge is important but that it is primarily accessed through professional networks and training programs instead of directly from peer-reviewed journals. Social media platforms seemed to be insignificant sources for professional knowledge. These insights raise important questions about how social workers use social media and the role of occupational networks and associations for the dissemination of research findings. Finally, our findings suggest that agencies and researchers think more purposefully about the infusion of knowledge into practice through opportunities for professional socialization, the use of research briefs, and open-access, peer-reviewed journals.

important to reiterate the inconsistencies between formal definitions of the concept of EBP and its actual implementation in the field. The original definitions conceptualize EBP as a process that includes asking empirically answerable questions, finding and evaluating the best available evidence, and applying that evidence in conjunction with client characteristics and practitioner judgment (Evidence-Based Medicine Working Group, 1992). The actual implementation in the field, by contrast, seems to draw on a much narrower definition that emphasizes research evidence based on randomized clinical trials that control for idiosyncrasies of the very contexts in which everyday practice situations are embedded (Chonody & Teater, 2018;Drake et al., 2007;Zayas et al., 2011). Particularly the omission of practitioner expertise in the adaptation of the EBP model led to intense debates about what constitutes appropriate knowledge and research evidence for social work practice (see Allen-Meares & Lane, 1990;Botha, 2012;Fook, 2001;Gilgun & Abrams, 2002;Hartmann, 1990;Imre, 1991;Osmond & O'Connor, 2004;Parton, 2000;Rosen, Proctor, & Staudt, 1999;Sheppard, 1995;Trevithick, 2008;Webb, 2001). Regardless of where one may stand on these arguments, the failure to adequately recognize the central importance of practice knowledge and the particularities of client circumstances as a critical component of the EBP process created a seeming disconnect between formal empirical knowledge and informal practice knowledge in social work (Herie & Martin, 2002). Carelse and Dykes (2014), for instance, found that social work students experience this tension between different kinds of knowing as a gap between their theoretical and formal coursework that tends to promote formal knowledge and their real practice experiences in the field.
The introduction of translational science can be viewed as the most recent effort to provide a bridge between social work research and practice (Brekke, Ell, & Palinkas, 2007;Brownson, Colditz, & Proctor, 2017;Hudgins & Allen-Meares, 2000;Palinkas & Soydan, 2012) particularly in light of the rapid expansion of social work research evidence and the concurrent reduction of barriers to access this information through free article repositories (Howard, McMillen, & Pollio, 2003). A recent study by Pendell (2018), however, found only marginal use of such repositories for sharing research published in the top 25 social work journals.
It is important to note that while translational science tends to examine variables and contextual influences on knowledge use in professional practice, EBP refers to the actual application of evidence in practice (Hudgins & Allen-Meares, 2000;Titler, 2018). Despite this distinction, translational science raises important questions about the effectiveness of dissemination of research findings through academic journals or the accessibility of scientific language that can pose serious obstacles to the conversion of this knowledge into practice. The need for research-practitioner partnerships to generate evidence that is useful for professional decision-making (Proctor, 2003) indicates, however, that notions of "bench-to-trench" knowledge production can reinforce tensions between formal and informal knowledge in social work. While the "bench" signifies academic or laboratory settings as the source of scientific or formal knowledge, the "trench" indicates practice fields as the origin of informal knowledge or practice wisdom (Fook, 2004;Proctor, 2003). There certainly exists a need to develop a deeper understanding of how to better implement the growing body of social work research evidence (Brownson et al., 2017). In addition, there is also a dearth of research on the actual sources and content of knowledge that practitioners draw on when making practice decisions. Research in kindred occupational fields such as education or nursing, however, has begun to examine the role of "non-bench" sites such as online communities and social media as sources of professional knowledge and development (Duncan-Howell, 2010;Moorley & Chinn, 2015). A recent study of teachers' sources of practice knowledge, for example, found an increasing use of Pinterest as a source for the development of pedagogical content knowledge (Grote-Garcia & Vasinda, 2014). While social media is becoming a critical tool of communication for political actors, social movements, professional associations and non-profits (Guo & Saxton, 2014;Margetts, John, Hale, & Yasseri, 2015;Segerberg & Bennett, 2011), there is currently a lack of research on the role of online platforms such as Pinterest, Facebook, or Twitter for the dissemination of knowledge and information that influences social work practice.
Aside from the sources of this knowledge, research indicates that academic or formal knowledge may not be the primary content to guide social work practice, but that instead practitioners tend to privilege information that helps them solve everyday practice problems. A survey among NASW members, for instance, showed that the types of knowledge considered most useful for practice was on particular social problems or clinical diagnoses, followed by assessments of the effectiveness of interventions, and information on specific client populations (Cha, Kuo, & Marsh, 2006;Marsh, Cha, & Kuo, 2004). A more recent study of German social workers' attitudes towards EBP by James, Lampe, Behnken and Schulz (2018) revealed a general openness toward research-based knowledge. The study findings, however, also indicate that these kinds of formal knowledge were not the primary sources to guide practice decisions but that social workers mostly relied on less formal types of knowledge such as practice experience and collegial advice (James et al., 2018). Beyond the recognition that utility for practice determines whether knowledge is used by social workers, we know very little about the specific content and sources of this type of knowledge.
This paper seeks to explore the following questions: (a) What are the most important sources and content of knowledge used by social work practitioners? and (b) How did this knowledge change professional practice? A better understanding of the sources and types of knowledge that practitioners emulate is an important contribution to efforts to close the practice-research gap in our field. We present findings from a survey of social work field instructors in St. Louis, Missouri. The analysis is based on their responses to mostly openended questions exploring access and use, as well as types, of knowledge that practitioners draw on in their professional practice. The findings highlight patterns and themes identified by practitioners that provide important implications for future research as well as different modes of disseminating its findings in social work.

Methods and Data
The analysis for this paper draws on data from a Qualtrics survey of social work field instructors in the St. Louis metropolitan area. The survey comprised a mix of mostly openended and some closed-ended questions that covered three main topics: (1) respondent professional background, including field and length of practice, current position, and type of agency (publicprivate), (2) types of knowledge that recently influenced professional practice, as well as sources and content of that knowledge, and (3) needed program content for continuing education.
The survey instrument consisted of 17 items: 7 closed-ended questions and 10 openended questions. Closed-ended items included questions about educational background, years of professional practice, and type of agency employer. Open-ended items were primarily aimed at exploring sources, content, and impact of influential knowledge for practice. The instrument was developed by the authors (an LCSW, and the director of the MSW program at a local university), and was tested internally among MSW students and faculty members to adjust for wording and content of questions, as well as the flow of the overall instrument.
The survey was sent via email to all 516 MSW field instructors in a joint database used by the field instruction offices of all three local social work graduate programs. The survey was live for about 4 weeks from mid May 2018 to early June 2018 and generated 349 responses yielding a response rate of 67.6 percent. Open-ended responses were coded into categories of knowledge sources (e.g., professional website or continuing education). The analysis presented in this paper is based on 250 completed surveys. Field instructors were chosen as the sampling frame because of their potentially multiplying influence on future social work practitioners in terms of their use of various sources of knowledge that shapes professional practice in social work.

Sample Characteristics
About 57% of our respondents were employed in a public agency and 43% in a private agency. Over 90% listed an MSW as their highest degree, while 2.4% had a BSW and 1.2% had a doctorate. About 6% of our respondents had a different degree (MPH, MBA or joint degree with MSW) and more than 73% had a professional license. Almost all of our respondents had at least 2 years of professional social work experience. The majority of respondents had between 5 and 20 years of experience and almost 32% reported more than 20 years of experience in the field. This is not surprising given that our sample consisted of social work field instructors. It is important to note that the questions about respondents' characteristics were limited to information deemed directly relevant to an exploration of sources and content of knowledge that influences social work practice, such as professional background and access to sources of knowledge. Information on gender, race, and age (although years of professional practice may serve as an index for age) by contrast, did not seem directly relevant for shaping sources and types of influential knowledge. Figure 1 displays the results for the sources of professional practice knowledge respondents have access to through their agency. Respondents were asked to list all relevant sources. All respondents identified at least one source for professional knowledge and the most prevalent were Continuing Education (19.7%), and events and co-workers (18.5%). Professional conferences (13.8%) and professional websites (12.3%) were also regularly mentioned. Future research will need to explore the relative importance of professional conferences as an opportunity to connect with other colleagues as opposed to the more content-based workshops, panels, and presentations.

Figure 1. Access to Sources of Knowledge
In light of the growing importance of online platforms such as Pinterest as a professional resource for teaching professionals (Grote-Garcia & Vasinda, 2014), it was somewhat surprising that social media (9.9%) did not feature more prominently as a source of professional knowledge among the social workers in our study. Finally, academic journals (5.2%) are only slightly more accessed than newspapers (4.9%) as sources for professional knowledge provided through their workplace. Considering the cost of access to academic journals and the slow uptake of free article repositories among social work journals (Pendell, 2018), this is not surprising.
The insignificance of social media raised questions about whether these responses differed depending on years of professional experience (as indicative of age) and field of practice. Table 1 shows the summary results of our cross-tab analysis that breaks down access to sources of knowledge by respondents' years of professional experience.
Respondents were asked about the sources of knowledge provided by their agency and could select all that applied from a list of responses. The results in Table 1 are not to be read as rank-ordering of knowledge sources but rather as a quick orientation to what sources were most often provided by the agency depending on the respondents' years of professional practice. Furthermore, years of practice are non-exclusive categories. We acknowledge that this is a measurement error but since this table represents a magnitude of themes rather than an accurate count we argue that a more precise scale would not alter the findings.
Since there were only two respondents in our study who had two years or less of professional experience, we omitted the data for these two respondents. There are three main impressions that emerge from this analysis: (a) the overall pattern of access to sources Academic Journal Newspaper of professional information is similar across the different experience levels, (b) regardless of length in the field, respondents highlight the critical importance of continuing education and co-workers as their primary sources for professional knowledge, closely followed by professional conferences, and (c) newspapers, professional journals, and academic journals, by contrast, were consistently ranked as the least important sources for professional information.
Furthermore, and contrary to our expectation, social media's importance as a source for professional knowledge and information was considered slightly more important for social work field instructors with 5 or more years of professional experience. This finding raises interesting questions for future research about the use of social media in the context of professional social work practice. One plausible explanation would be that social workers with fewer years of professional experience view social media less as a professional tool compared to those with more years of experience. We then filtered all respondents who were in macro practice positions, such as managers, administrators, directors, and team leaders (n = 114) in order to examine potential differences in their access to sources of knowledge. Figure 2 shows that social work administrators in our sample did not differ in the types of sources of knowledge their agencies provided.
Next, we wanted to learn more about the types of knowledge that most recently changed our respondents' professional practice, the sources and content of that knowledge, and how it changed their practice. These questions were open-ended and we coded the responses into categories of types, sources, and content of knowledge.

Sources of Knowledge That Change Practice
The most common source of knowledge that recently had an effect on practice among our respondents was some form of professional training or continuing education. These programs included university-based programs, coaching schools, agency programs, or field instructor trainings. The importance of formal continuing education programs as sources of professional knowledge may in large part be due to the requirements of CEU for license renewal. The next category of sources for professional knowledge were professional journals, academic publications, conferences or associations, co-workers, and professional networks. These results are also consistent with the types of sources our respondents have access to through their agencies and they emphasize the importance of professional networks and conferences.
While academic journals were ranked close to the bottom in terms of their accessibility, it is interesting to note that among administrators, along with expert presentations (e.g., researchers and consultants), academic journals and research reports fared noticeably better as sources for knowledge that influences practice. This could indicate one pathway through which research articles and peer-reviewed research may still find their way into professional networks and continuing training programs, despite the limitations to direct access of academic journals. The importance of academic and expert sources was closely followed by government websites, including the Centers for Medicare and Medicaid Services, U.S. Department of Housing and Urban Development (HUD), Centers for Disease Control and Prevention (CDC) and the Department of Education, as well as online sources such as webinars, podcasts, and videos. And finally, newspapers and social media, including Facebook and Twitter, were least often noted as sources for information or knowledge that had recently had an effect on professional practice. We then compared the sources of influential knowledge between administrators and practitioners in our study. Table 2 shows that continuing education and professional training courses are the most common sources of influential knowledge for both groups, but that professional journals, conferences, or associations as well as online resources such as webinars, podcasts and TED talks are much more prevalent among practitioners, while expert presentations and co-workers are more often named as knowledge sources among administrators. Academic journals are much less prevalent sources for influential knowledge among practitioners compared to administrators. These findings of sources for influential professional knowledge raise interesting questions about the relative importance of different means, such as continuing education programs, professional networks or online platforms for the diffusion of information and knowledge in our field. But aside from the sources of such knowledge we also wanted to know what the content of this knowledge was and how our respondents thought it had changed their practice.

Knowledge Content
The most prominent types of knowledge that turned out to have influenced the professional practice of our respondents could be summarized under the general category of "what works." This is consistent with previous research on knowledge use in social work (see Cha et al., 2006;James et al., 2018;Marsh et al., 2004). This category included training for, and background research on, new interventions and treatment models such as training on signs of safety in child welfare, trauma and different modes of trauma treatment, moral reconation therapy or eye movement desensitization and reprocessing (EMDR). Knowledge and information about the effectiveness of treatment models and interventions was the second most-often named type of knowledge that had influenced practice among our respondents.
Aside from training and information on treatments, information on policy and program changes, mostly derived from government and agency websites were also frequently considered influential forms of knowledge that changed our respondents' practice. This type of knowledge includes information on changes in Medicare and Medicaid coverage policies, McKinney Vento Act updates, or changes in the Missouri Criminal Code. Information on policy changes was closely followed by knowledge content on supervision and administrative techniques such as leadership training, different processes for agency data collection, information on caseload sizes in child welfare, or development of staff capacity. A few respondents also mentioned information on occupational licensure and professional conduct as helpful.

Effects on Practice
How did this knowledge change or influence our respondents' practice? Nearly half of respondents (49.6%) noted the acquisition of additional professional interventions or tools and the expansion of services provided as the most frequent effects on their practice. These changes included the restructuring of the care management department at an agency, adjustments in how an agency helps foster families respond to behavioral problems of children in their care, additional therapy treatments to better support clients, and different approaches to structuring supervision sessions. In addition to the development of professional skills and services, changes in perspective on their practice (34.8%) was the second most prominent way in which new knowledge facilitated shifts in professional practice. These perspectives ranged from using a trauma-informed lens to examining interventions and services, acknowledging a client's right to be wrong, rethinking abstinence as an appropriate objective for substance use treatments, and increasing awareness of racial dynamics in agency interactions. For others (8.2%) the new knowledge provided evidence supporting current agency practice or offered a refinement and extension of earlier practice. And finally, some respondents also mentioned that the new information helped them reflect on personal issues (7.4%) such as difficulties with listening and empathy, avoiding discomfort as a barrier to personal and professional growth, and realizing a better work-life balance.

Discussion
This study explored the sources and content of knowledge that influences social work practice from the vantage point of field instructors. The analysis showed that colleagues and professional development programs were considered the most accessible sources for influential knowledge and information. This is not surprising given the expense for online access to research publications. It was interesting to note that social media was ranked in the middle range as a source for professional knowledge and this assessment did not vary much across respondents with different years of professional experience. Access to various types of sources of information certainly influenced responses about the origins of the most recent knowledge that had an impact on our respondents' professional practice. Professional conferences, peers, and continuing education programs were among the most often noted sources of such knowledge. This makes sense in light of the requirements for professional license renewal. But in contrast to the relative difficulty in accessing academic journals, expert presentationsincluding researchersalong with research reports and academic publications, ranked in the middle range of sources for influential knowledge. In addition to professional networks and events or experts, a set of government or foundation websites were most commonly noted sources for information, primarily with regard to changes in policies and regulations. Very few respondents reported receiving influential information through social media platforms such as Facebook or from newspapers.
The most frequently named types of knowledge content were training in new interventions but also research on background and effectiveness of treatments. This confirms findings from previous studies that indicate that social workers are most interested in information on what works (see Cha et al., 2006;James et al, 2018;Marsh et al., 2004). But these findings also highlight that, while research knowledge has a critical influence on practice, it is not necessarily accessed through peer-reviewed journals despite growing proliferation of open access sources for academic publications. Instead, the results of this study point to the critical importance of occupational peer networks, professional conferences, and continuing education programs as a critical means for the dissemination of research evidence among social work practitioners. While theories of social networks or social capital note the central importance of relationships and connections with others for accessing critical information and support (Burt, 2009;Coleman, 1988;Portes, 1998), their role for the dissemination of professional knowledge in social work is currently not very well understood. For instance, it would be helpful to learn whether the importance of professional conferences for knowledge dissemination is primarily due to their formal and expert presentations or whether they are primarily an opportunity to socialize with professional peers. In addition, our study did not elaborate on whether professional training and conference participation was facilitated by the respondents' employer or whether they were sought out by the respondent, which may be a potentially relevant aspect of accessibility of professional knowledge.
Although some respondents noted online resources such as webinars, videos, podcasts and TED talks as sources for new knowledge, the relative insignificance of social media was somewhat unexpected. Particularly the latter finding raises questions about the role of social media platforms such as Facebook or Twitter for the distribution of influential knowledge in our profession.
Aside from sources and content, our analysis also revealed that the influence of new knowledge on professional practice operated primarily on a technological and a reflexive dimension, either by expanding intervention tools and services provided, or by changing the perspective of social workers on their practice. It is important to reiterate that these findings are based on a localized sample of field instructors and do not capture sources and content of knowledge used by all social workers in St. Louis or elsewhere. Furthermore, in a conscious effort to keep the number of items on our survey at a minimum in order to be considerate of the busy schedule of social workers and to increase the response rate, we were not able to probe for the role of agencies in selecting training content or whether conferences are an opportunity for professional networking. And finally, while the analysis shows differences in sources of influential knowledge between administrators and practitioners, it did not explore these differences for licensed and non-licensed workers. This could be an interesting question for future research. Given these limitations, however, the findings on the patterns of knowledge use among seasoned social work field instructors raise important questions about the sources and content of knowledge that changes practice in the field, the role of professional peer socialization, as well as social media and other online sources for the dissemination of knowledge in the social work profession.
More importantly however, these questions urge social work researchers and agencies to think more purposefully about effective mechanisms to infuse knowledge into the practice of social workers in the field. For example, in addition to agency administrators deciding on appropriate topics for professional training sessions, it may be beneficial to think about providing opportunities for formal as well as informal professional socialization with colleagues through inter-agency workshops or at professional conferences. Although social media did not feature prominently as a source for professional knowledge, online platforms or formats, ranging from periodic research briefs compiled by social work researchers to more open-access academic journals would greatly increase access to research knowledge. This study also raises a question about the potential to enhance professional practice through trainings on how to leverage new technologies and social media. After all, while we agree with Einstein's statement that experience is important, it is not the only source of knowledge.