Behind the Wall: An Exploration of Public Access to Behind the Wall: An Exploration of Public Access to Research Articles in Social Work Journals Research Articles in Social Work Journals

: Despite implicit and explicit expectations that research inform their practice, social workers are unlikely to have access to published research articles. The traditional publishing model does not support public access (i.e., no publisher paywall barrier) to scholarly journals. Newer models of publishing allow free access to research including open access publishing and deposit of scholarship in institutional or disciplinary repositories. This study examined public access to articles in the top 25 social work journals. A random sample of article citations from a total of 1,587 was assessed, with the result that 52% of citations had no full-text access. Of the remaining 48% of citations with full-text access, it is questionable most will remain available long term due to possible copyright violations. Citations from the random sample show only minimal usage of institutional or disciplinary repositories as a means of sharing research. Establishing this baseline measure of access to research is an important first step in understanding the barriers for social workers in accessing research to inform practice. Recommendations for increasing access to research include publishing in open access journals and utilizing full text repositories.

As a social work librarian, I have many conversations with students regarding access to research articles.When students get close to graduation they often inquire if and how they might still use the library's databases and full-text journal access.Unfortunately, my answer is always disappointing.After graduation, their off-campus access to our electronic resources ends, and their only remaining means of access is to come into the Library to use public access computers or connect to campus guest wi-fi, which of course hardly anyone has time to do with busy work days and other obligations.Graduating social work students already know from their field placements that access to full-text journals is not often available from agencies either.In a focus group study on lifelong learning by Jivanjee, Pendell, Nissan, and Goodluck (2016), one MSW student expressed her concern: I love libraries, but I know that my access to this library in particular is going to be next to nil once I graduate… There isn't much out there for those of us that like research and want to base it [practice] on research.We are a little bit in a hole.(p. 267) The Dean of the School of Social Work at my institution has informed me that she continually receives requests from alumni for off-campus access to databases and full-text journals.This access is seen as one of the most useful benefits they could receive.Aside from stripped down alumni packages from vendors, academic publisher and other vendor licenses prevent the Library from extending access to individuals not actively affiliated with the University.It is commonly recognized by librarians that even if licenses were to permit access equivalent to that of affiliated users, the cost and management of providing the level of access desired would be enormous and complicated.Anecdotally, many fellow social work librarians are familiar with this dilemma of graduating students and alumni requesting services and resources that cannot be provided.
That students and alumni expect, or at least hope, for access to research literature is understandable considering their recent education and the professional standards laid out for them.For example, the National Association of Social Workers (NASW, 2017)  Additionally, evidence-based practice (EBP), or similar research-informed practices in the field, overtly expect, or at least imply, that social workers have meaningful access to research relevant to their practice needs.
It would also seem safe to assume that researchers want their efforts to influence practice in the field and improve the lives of the communities with whom they work.The NASW has expressed interest in connecting research to practice.The NASW Action Network for Social Work Education and Research (n.d.) states: "Social work research provides empirical support for best practice approaches to improve service delivery and public policies."Additionally, an NASW (n.d.) advocacy briefing on social work research offers among its recommendations: "Promote the dissemination and implementation of research into real world settings and to encourage communication between researchers and practitioners" (p.2).
Observing the tension between production of research for the field and the field's lack of access to research, this study attempts to answer an initial, primary question: if a practicing professional wants to read research literature, is the research literature available to them?Ten years ago or more, the answer would likely have been "no," as the traditional, subscription-based publishing model held content more firmly behind paywalls.The traditional model of publishing severely limits the public's access to research literature, despite the content and peer review expertise provided by researchers, most often gratis (see Figure 1).However, new pathways to research articles have been created and cultivated, particularly open access (OA) publishing and use of disciplinary or institutional repositories (IRs; see Figure 2).Articles published in OA journals, such as Advances in Social Work and Behavior and Social Issues are referred to as "gold" OA meaning that access is provided directly via the publisher.Access provided via institutional or disciplinary repositories is referred to as "green" OA, and the articles are usually in manuscript form, either prior to peer review or post peer review (referred to as pre-print and post-print respectively).The gold OA publishing model removes the cost burden of access from the public, replacing financial support for journals via a variety of models, such as subsidies, donations, or membership dues for professional association journals.Repositories also provide public access to research, but a bit differently; the ability to deposit an article might be subject to an embargo period, or, at times, only the pre or post-print version may be deposited, not the publisher's final formatted version.In the early 2000's, IRs were developed as universities leveraged new technology to create stable access to the research output of their communities.Adding public access to this effort was a natural fit.Additionally, disciplinary repositories were on the rise, particularly in the hard sciences where sharing research quickly is highly valued.Both IRs and disciplinary repositories serve as a primary means for sharing publicly funded research; for example, PubMedCentral (PMC), a health sciences research repository, has grown rapidly since the implementation of the National Institutes of Health public access policy of 2008 (National Institutes of Health, 2016).

Figure 2. Open Access Publishing Cycle for Research Articles
This study examines public access (i.e., no paywall barrier) to articles in the top 25 social work related journals.While none of the journals in the sample set are published OA, they each allow the author to archive a version of the article in an institutional or disciplinary repository.Measuring the amount of engagement in self-archiving and availability of research in social work is an important first step in advocating for practitioner and public access.
Quantifying the societal or practice-related impact of freely accessible research literature is challenging; however, it is possible to identify benefits of OA indirectly, especially through the lens of inefficiencies.In a study of government officials' use of research, performed early in the development of OA publishing, Willinsky (2003) found that: The research that is most easily accessible, through portals and open-access sites, is most often consulted, as policymakers referred to how readily they were dissuaded from using pay-per-view and subscription services in their pursuit of knowledge.This means that they are tapping into a skewed and somewhat haphazard view of the current state of knowledge on a given topic.(p.12) Unfortunately, almost ten years later, Look and Marsh (2012) found similar results in their study of public sector employees.For example, the participants "noted that they submitted analysis or made decisions on the basis of potentially incomplete research as they had not been able to review all the relevant articles they had identified" (p.29) or that decisions were delayed due to lack of access.
Given the demands on social workers to stay up-to-date in their practice, use research to inform decisions, and to participate in continuing education and lifelong learning, easy access to research literature can only be a benefit.But, the traditional scholarly publishing model is still focused on providing research to other researchers, and not extending into the practitioner and public audiences.As an example of a different approach to the dissemination of scholarship, Advances in Social Work (AISW), one of the few social work OA journals, states in its Open Access Policy "This journal provides immediate open access to its content on the principle that making research freely available to the public supports a greater global exchange of knowledge" (AISW, n.d.).

Literature Review
Aside from the earlier recommendation of promoting dissemination of research by NASW, there appears to be little, if any, conversation in social work related journals regarding access to research for practitioners.Several keyword and related subject heading searches for open access or dissemination of research in Social Services Abstracts, Google Scholar, and Web of Science returned very few relevant results.Bowen, Mattaini, and De Groote (2013) was the only study found that explicitly addresses access to research literature and open access publishing.
Another, albeit brief, exception to the quiet was "Suggestions to Improve Social Work Journal Editorial and Peer-Review Processes: The San Antonio Response to the Miami Statement" (Holden et al., 2008) which offers fifty responses from a subcommittee of the Society for Social Work and Research Presidential Task Force to an earlier "Miami" statement, which was also concerned with improving journal publication processes.Both of these statements centered on concerns regarding scholarly communication via social work journals (Holden et al., 2008;Schilling et al., 2005).Recommendation 48 of the "San Antonio" response states: "Publishers should consider ways in which they can increase the free flow of scholarly communication (e.g., removing restrictive practices regarding the use of copyrighted materials, becoming a green journal, and removing publication process obstacles in the production realm)" (Holden et al., 2008, p. 69).
The primary source of any data or discussion regarding access to research literature in social work is primarily found in studies related to EBP because a key component of EBP is finding research that matches one's particular clinical or practice question.Often, the expectation that one can access the needed research is implicit, and perhaps taken for granted.But studies looking specifically at EBP in the field of social work have demonstrated that needed information access is not typically available: Seven of the studies identified poor access to available research evidence as a barrier to EBP implementation.The need to invest resources in staffed library facilities and information technology to access web-based databases was identified as a requirement if there were to be a movement from EBP as an aspiration to a reality.(Gray, Joy, Plath, & Webb, 2013, p. 163) Lack of access to information resources was also cited as one of the top three barriers to EBP in a study recent of Romanian social workers (Iovu, Goian, & Runcan, 2015).
It is interesting to observe dramatic differences in engagement with EBP between practitioners in the field and students.Van der Zwet, Weling, Beneken genaamd Kolmer, and Schalk (2017) explored if social workers simultaneously enrolled in an MSW program were more engaged with EBP than their non-enrolled counterparts.Indeed, the difference appeared substantial; only 12.3% of social workers reported that they "use the Internet to search for the best research evidence to guide my practice decisions," in contrast to almost 60% of social worker/MSW students.The study does not further clarify or explore what type of information found on the Internet is being used; however, articles seem likely to be part of this scope.Additionally, 75% of MSW students "often" or "very often" read research evidence as part of their EBP process, again in contrast to only 10.6% of practitioners ( Van der Zwet et al., 2017, p. 84).A question that arises from this study is to what extent access plays a role in identification and use of research evidence.It can be assumed that the MSW students in the study have full-text journal access via their university affiliation, while the practitioners do not.Would we observe more use of research if practitioners had the same access as students?Hardisty and Haaga (2008) attempted to answer this question by exploring the impact of access to articles to influence behavior/practice change.Their unique study explored participants' behavior when asked to access an article that was freely available versus other means of access, including needing to pay for the article.A significantly higher proportion of participants read the article when it was freely available to them (70%; p. 831).The same authors also tested for the potential of the participants to implement recommendations from the article in a relevant hypothetical scenario, finding a positive result."Dissemination efforts that come with a price tag may prove less effective, even if they are promoted more heavily, than dissemination efforts that are free" (Hardisty & Haaga, 2008, p. 835).They conclude that "A more immediate practical implication of the present study is that scholars wishing to maximize the diffusion of their research among the professional community should deposit eprints of their work in OA archives" (p.836).
If the evidence points to access as a positive benefit to practitioners, why is more access not provided to practitioners via subscriptions?The cost barrier to academic journals is high for individuals and agencies.Bowen and colleagues' (2013) study of social work journal subscription cost data found an average of $121 per subscription for an individual, with much higher costs for institutions.The average yearly inflation rate for academic journals tends to hover in the 5%-6% range, but the projected increase for social sciences titles for 2017 was 7.2% (Bosch & Henderson, 2017).Bowen et al. (2013) reasonably assert that "open access is particularly important when considering the needs of practitioners in less wealthy organizations, communities, and institutions in the United States and globally," (p.40) especially when promoting evidence-based practice.
Acknowledging that the high impact journals in the field of social work do not include any open access titles, there is another key pathway for researchers to share their work: institutional or disciplinary repositories.Institutional repositories are the most viable path for stable, long term, and publicly/freely available sharing of research, outside of publishing in an OA journal, and IRs are common at all major universities at this point.In 2017, there were 820 institutional repositories in the United States, Canada, and the United Kingdom according to the Registry of Open Access Repositories.Disciplinary repositories, such as SocArXiv, provide an open access platform for researchers to upload pre-prints and working papers; repositories such as PubMed Central provide a platform for published papers.Overall, social sciences disciplinary repositories are younger and smaller than those in the hard sciences, but they have seen growth in their collections.PubMed Central and its counterpart, Europe PMC, have grown significantly, assisted by being the recommended depositories for health sciences research subject to public access policies.
Many IRs are supported by the university library, with varying levels of promotion and support for depositing content.The rate of self-archiving in 2010 United Kingdom was 34% for the social sciences broadly, and psychology more specifically (Van Noorden, 2012, p. 303).However, the United Kingdom also has more robust public access policies in place than the United States.Overall though, the rate of self-archiving of research articles by authors lags far behind its potential, especially when compared with the swift enthusiasm a publication sharing platform such as ResearchGate appears to generate.

Method
In order to measure the level of public access to research articles in the field of social work, citation searches were performed on a sample set of articles from the top 25 social work related journals.Using the Eigenfactor metric of article influence for 2014, the top 25 journals in the International Scientific Indexing category of social work were identified (see Table 1).Eigenfactor's article influence score measures the average number of citations to individual articles within a journal; this measure was selected as article level access was the primary concern for this study.Ultimately, the bulk of the top 25 journal titles with high article influence are the same as those ranked by overall journal influence.
I retrieved citations for articles published in 2014 from each of the journals using the Social Sciences Index available via the Web of Science.As part of this process every attempt was made to accurately exclude citations for editorials, book reviews, letters, corrections, or other citations that were not recognizable as research articles.None of the 25 journals identified are open access; however, all of the journals allow for the deposit of a pre-print or post-print version in an institutional or disciplinary repository.All but five titles allow for paid open access of individual articles, as indicated by SHERPA/RoMEO, an aggregator of journal OA policies (SHERPA/RoMEO, n.d.; see Table 1).The citation sample set was derived from citations from 2014, allowing for an over two year gap to account for the embargo period sometimes required by publishers prior to self-archiving in a repository.From the total number of 1,587 research article citations retrieved from the 25 journals, a random sample of 638 citations created by a random generator for Google Sheets was selected in accordance with a 95% confidence level with +/-3 confidence interval.
Each citation was searched using the Google Scholar and DOAI.io in an off-campus, unauthenticated browser to avoid inaccurate access to full-text via IP recognition or institutional affiliation.Google Scholar is a well-known discovery and access point for scholarly journal articles; its search engine crawls academic publisher websites as well as institutional and disciplinary repositories, such as PMC.Google Scholar also crawls ResearchGate for full-text articles uploaded by its users.DOAI.io is the Digital Open Access Identifier, which utilizes an article's Digital Object Identifier (DOI) to retrieve free access to a version of the article if available.Each citation was searched using both applications because they demonstrate different access points for a given citation.Neither of these discovery tools search full-text "pirate" sites such as SciHub; access to full-text via those sites is not part of this study.
A pilot was performed with 20 citations, revealing that in addition to the planned recording of availability of full-text from an institutional or disciplinary repository, download availability of full-text via the social networking website ResearchGate should also be noted.Therefore, the public availability, if any, of full-text (publisher PDF, postprint, or pre-print version) via a DOI search of Google Scholar and DOIA.io was recorded, including the source of availability as follows: • Repository (institutional or disciplinary) • ResearchGate • Open access on journal website • Other full-text access (e.g., personal or organizational website)

Findings and Discussion
As seen in Table 2, no full-text access was found for 52% of articles, a slim majority of the total 638 citations investigated.The sample set of citations demonstrated only minimal use of institutional repositories as a means of sharing research.Deposit of fulltext in the disciplinary repositories PMC and Europe PMC was higher than that of institutional repositories, and likely can be attributed to funding agency or governmental requirements, as discussed earlier.Within the parameters of this study, very few articles would be available to users not affiliated with an institution or some other form of subscription access without the popularity of ResearchGate, which accounted for 30% of full-text access.In regards to the high use of ResearchGate, it is likely that many articles uploaded to the platform are in violation of publisher copyright agreements.In fact, mass take down notices from publishers targeting ResearchGate have already begun (Van Noorden, 2017).As Jamali (2017) notes, the number of recent articles on ResearchGate is "...not good for publishers as journals' main revenue relies on newly published issues and that is why a lot of publishers apply embargo periods for archiving peer-reviewed versions of the articles" (p.252).Jamali (2017) found over 78% of a random sample of 500 articles in ResearchGate were publisher PDFs (p.251).This study's results demonstrate an even higher rate of deposit of publisher PDFs of the articles found in ResearchGate: 96% of articles on ResearchGate were publisher's PDF version.
Ascertaining the means by which these articles were uploaded to ResearchGate, or the amount of them that are in violation of copyright is outside of the scope of this study.However, the legal complications of ResearchGate are important for our understanding of public access because this platform appears to be a primary means of access that would be otherwise unavailable."To understand the role of ResearchGate in making full-text of papers freely available, it is enough to say that it is one of the top sources of full-text files found through Google Scholar" (Jamali, 2017, p. 242).This will change quickly as publishers take legal steps to limit the extent of copyright violations via ResearchGate (Van Noorden, 2017).Here the distinctions between IRs and disciplinary repositories versus forprofit social network platforms becomes clearer: repositories are born with a different purpose and ethos, and are managed with an eye towards copyright compliance and long term access.
As a simple tracking of full-text availability, this study does not answer the question of why engagement in self-archiving is low.For example, the availability of institutional repositories to authors in the sample set and the amount of support offered by the institution is unknown.Authors may struggle with self-archiving in repositories due to the "time and effort involved in determining or securing copyright" (Palmer, Teffeau, & Newton, 2008, p. 25).Or perhaps the low rate of deposit is due to lack of awareness and education regarding the purpose and utilization of OA, IRs, and disciplinary repositories.Further research related to both the researcher/author perspective and the practitioner/user perspectives is needed.It would be valuable to have a better understanding of whether researchers perceive the impact of their work primarily as citations from other researchers or for use by practitioners in the field.The level of awareness and understanding of options for sharing research among authors are also of interest.
Looking at practitioners/users, to what extent do they use, or wish to use, research articles to inform their practice?Are the aforementioned pirate sites or other avenues (e.g.#canihazpdf on Twitter) providing viable work-around access for savvy social workers?And for social work librarians, we must ask if teaching students to use expensive, licensed databases that lead to subscription-based full-text platforms is effective for students once they separate from the university.Would it be helpful to discuss the economics of scholarly communication to increase understanding of how access is available or not available?Additional research on topics like these, building upon this study, has the potential to inform us how researchers/authors think of dissemination of their work; how students are taught to find and access research; how access shapes decision-making in the field; and how it influences OA-related advocacy efforts in social work.

Conclusion
The findings of this study establish a baseline measure of research article availability to practitioners and the public, assuming no personal and institutional subscriptions to individual journal titles.A significant amount of research published in core social work journals is not available to the practitioners or members of the public who would benefit from it.This is especially of concern as public skepticism of academia and scientific research increases.It is time for researchers, administrators, publishers, and librarians in the field of social work to increase openness and access to research.Looking ahead, there are two immediate ways for researchers to move their scholarship from behind the publisher paywall: publish in an OA journal whenever possible and practice self-archiving in either an institutional or disciplinary repository.Many will likely have the support of their university library in these efforts.Clearly, there is at least some interest in sharing scholarship as demonstrated by the use of ResearchGate, but as noted, there are caveats to such platforms.Librarians have a role in educating both researchers and students about the differing models of scholarly publishing, author's rights, and public access indexing tools and OA full-text sources.A longer term recommendation for researchers and administrators in an academic settings to consider is rewarding OA publishing and/or repository archiving as a reflection of community engagement in processes such as promotion and tenure.In addition to the societal benefits of OA publishing, a majority of comparison studies have indicated that OA publishing also benefits authors by increasing citations to their articles versus non-OA published articles (Tennant et al., 2016).Gold and green OA come with their own concerns and complications to be sorted out.However, the status quo of keeping research behind publisher paywalls costs the field of social work in ways that should not remain unexamined.
Code of Ethics Standard 4.01 outlines the expectation that: (b) Social workers should strive to become and remain proficient in professional practice and the performance of professional functions.Social workers should critically examine and keep current with emerging knowledge relevant to social work.Social workers should routinely review the professional literature and participate in continuing education relevant to social work practice and social work ethics.(c) Social workers should base practice on recognized knowledge, including empirically-based knowledge, relevant to social work and social work ethics.

Figure 1 .
Figure 1.Traditional Publishing Cycle for Research Articles

Table 1 .
Journal Sources of Citations with Eigenfactors, Number of Article Citations, and RoMEo Status