Afﬁrming our commitment to research: the Medical Library Association’s research policy statement: the process and ﬁndings

Purpose: Building on its 1995 research policy statement, the Medical Library Association (MLA) has issued a new research policy, The Research Imperative . This paper shares the background research that informed the new policy. Methods: Semi-structured interviews were conducted with ﬁfty-one key informants representing various library types, functions, geographic locations, ages, and ethnicities. The grounded theory approach was used to analyze the resulting textual database. Additionally, to gather input from the membership as a whole, two open forums were held at MLA annual meetings. Results: Key informant data indicated that the policy should provide roles for MLA in leadership, advocacy, collaboration, services, education, publishing, and development of a research agenda. Evidence-based library and information practice was emphasized. Six themes emerged to center the new policy: creation of a research culture, challenges, domains of research, research skills set, roles of stakeholders, and measurement of progress. Conclusion: Reﬂecting the interests and beliefs of the membership, The Research Imperative challenges MLA members to build a supportive culture that values and contributes to a research base that is recognized as an essential tool for future practice.


INTRODUCTION
Early in 2005, Joanne Gard Marshall, FMLA, then president of the Medical Library Association (MLA), created a task force to review and revise Using Scientific Evidence to Improve Information Practice [1], the association's research policy statement published in 1995. The task force also was charged to restructure, enhance, and develop MLA programs to advance research in the profession. In viewing its charge, the Task Force on MLA's Research Policy Statement chose to interpret research broadly, that is, as the systematic, careful, organized, and deliberate process of building a body of knowledge. This process can involve the discovery, confirmation, or revision of facts, concepts, principles, theories, or patterns of relationships between defined variables.
After initially focusing on the state of research in the profession and barriers and incentives to its use and creation, the task force agreed that it was important to do its own exploratory, qualitative research before making any determination about a revision. Such a study would provide the task force with a broader perspective of the issues related to research in the MLA membership as well as assess the impact and * The Research Imperative policy statement and the background research reported here were supported by and undertaken on behalf of the Medical Library Association (MLA). The Research Imperative is copyrighted by MLA.
A supplemental appendix is available with the online version of this journal. continued relevance of the 1995 policy. The study quickly made it clear that a revision was needed.
A revised and refreshed research policy statement, The Research Imperative, was accepted by the MLA Board of Directors in May 2007 [2]. This paper describes the task force's approach and its research, to highlight the findings and the process so that it can be replicated and/or improved in the future.

BACKGROUND
Using Scientific Evidence to Improve Information Practice [1] was prescient in its vision of health sciences librarians contributing to evidence-based medicine, organizational knowledge management, and effective integration of content into clinical and academic information management systems through their involvement in and use of research. The policy, while rooted in many years of MLA research-related activities and discussion, was innovative. It did not simply present a research agenda or committee charge as other professional associations had, but rather developed a vision and made a commitment for the association [3].
Subsequently, other library and information science (LIS) associations responded to the need for a research base for the profession with a range of publications. In 2001, the Special Libraries Association (SLA) issued its research policy statement, Putting Our Knowledge to Work [4]. SLA's policy is noteworthy for focusing on building the knowledgebase for evidence-based library and information practice and encouraging its members to conduct their own research into librarianship questions and share their findings with others. The Amer-• The Research Imperative reaffirms the Medical Library Association's (MLA's) commitment to research and emphasizes the need to develop health sciences librarianship's own knowledgebase as a solid foundation for best practices. • Informants described a growing recognition that health sciences librarians should base their practice decisions on evidence just as the health care practitioners and researchers they serve should. • A research process that gathered data systematically from the membership guided the policy's development. • Members expect the association to monitor progress toward its research goals and report on it annually.

Implications
• The Research Imperative challenges the association and its members to build a culture of reflective practice in which the profession's evidence base is routinely used. • To advance the vision articulated in its research policy, MLA must lead by inculcating research throughout the association, identifying a research agenda, advocating for improved access to and support of library and information science research, fostering collaborations, and educating members to better use and conduct their own applied research. • MLA should leverage its Center of Research and Education (CORE) as a repository for member-generated research, tools, and datasets.
ican Association of Law Librarians (AALL) issued its research agenda in 2000 ''to promote exploration'' of the changes in the work and workplace of its members brought on by ''the Information Age.'' Research questions in six broad categories were proposed for study [5]. The American Library Association's (ALA's) longstanding Committee on Research and Statistics (CORS) is charged with stimulating research among the various ALA divisions and advising the ALA Office for Research and Statistics on data collection issues and research questions that should be pursued [6]. Internationally, the movement to recognize the importance of research to the profession has grown as well.
The Chartered Institute of Library and Information Professionals-the United Kingdom's leading professional organization for librarians, information specialists, and knowledge managers-recently issued a policy and strategy for library research that linked research with development as both are necessary to achieve positive change and advance knowledge transfer services [7]. As in all disciplines, LIS research is in part predicated on an essential understanding of research practice, design, and methodology. Understanding research methods, along with the ability to analyze and interpret research findings, is among the core set of knowledge and skills for health information professionals, as defined by MLA's educational policy [8].

METHODS
To gather multiple views about MLA's role and responsibilities related to research, the task force followed several approaches that drew on a variety of qualitative methods appropriate for answering focused questions. Ongoing communication and collaboration with the Task Force on MLA's Educational Policy Statement, which was in a parallel process of reviewing and revising the association's Platform for Change [9], assured exchange of ideas related to training in research methods. MLA annual meetings usually include at least two open forums in which timely issues are presented to meeting attendees for discussion and feedback. Open forums at the 2005 and 2006 meetings provided the venue for dialogs among the MLA membership and task force members. After both annual meetings, brief items describing the discussions appeared in MLA-FOCUS, the association's biweekly online newsletter. The items invited members who had not attended the meeting or open forums to comment on the issues and process.
In addition to reaching out to the Task Force on MLA's Educational Policy Statement and the membership as a whole, the Task Force on MLA's Research Policy Statement embarked on a more systematic qualitative research process. The purpose was to gather opinions about MLA's 1995 policy statement and about research issues in general from a cross section of the membership and the profession, nationally and internationally. The task force determined that key informant interviewing was the most efficient method of gathering meaningful data in a relatively short period of time.
The selection of key informants was a purposeful critical case sample [10] intended to achieve a data set that would be both information-rich and broadly representative of the membership. Task force members all participated in identifying potential key informants. Criteria for selecting informants included a working knowledge of research either through providing information to support evidence-based medicine or previous conduct of research in the field. The sample also was intended to represent a range of positions and roles from director to reference librarian. Consideration was given to informants' geographic location; library type (including hospital, academic, or other); and demographics such as gender, age, and ethnicity. While the task force did take into account position and demographic factors in selecting informants, the study's purpose and methodology required that informants possess extensive knowledge of the subject rather than be a sample representative of the whole population. MLA reports survey results only in the aggregate, so findings are not directly attributable to individuals or institutions.
Interviews were semi-structured. An interview guide was developed outlining general questions designed to elicit open-ended responses (Appendix online). After two task force members pretested and validated the questions with a small group of informants, the other members of the task force began interviewing informants either in person, by email, or by telephone. Ultimately, each task force member interviewed four to six informants. As the number of interviews grew, gaps in viewpoints were noted and additional informants were added to the list of interviewees. Interviewing continued until the task force believed it had achieved saturation as indicated by redundancy in responses. All task force members transcribed their interview notes, and these were compiled into one large dataset. While all members read the compilation, two task force members conducted preliminary reviews of the dataset. One member (McKibbon) analyzed it question by question for commonalities among responses; the other (Perry) reviewed the dataset for ideas that could be used to guide recommendations to the association's leadership. Finally, using the grounded theory approach to analyze textual data and derive hypotheses or themes [11,12], the task force as a whole analyzed the full dataset to uncover underlying concepts and relationships, reaching consensus through discussion on six overarching themes that ultimately were used to structure the final document.

Open forums
Both open forums at the MLA annual meetings were conducted jointly by the research and education policy task forces. Generally, thirty to fifty people attended each of the two forums. The first forum in 2005 attempted to look at the broader issues facing the profession. Participants were invited to envision the future of libraries, librarians, and MLA in 2020. The picture that emerged was a world in which globalization was a fact of life; knowledge was democratized; partnerships and collaboration across corporate, government, and academic sectors were common; libraries were digital; and librarians with subject expertise were integrated or distributed into the workflow of special user groups. In this world, MLA was seen as fostering relationships, certifying subject experts, promoting connections between education and practice, and facilitating the creation of a shared knowledgebase for the profession. These ideas were the starting point for both task forces as they grappled with their charges.
A year later, the second open forum was used to gather feedback on the directions taken by both task forces. Attendees strongly affirmed the research policy task force's focus on evidence-based library and information practice, the need for a research agenda, and an annual review so the policy would be a ''living document.'' They also endorsed the proposed emphasis on collaborative research and asked that the task force share and publish its findings from the key informant interviews.
A small number of comments were received as a result of the MLA-FOCUS items, echoing comments from the open forums.

Key informant interviews
Fifty-one key informant interviews, including the 2 used to pilot-test the interview guide, were conducted by task force members (Table 1). Most interviews were conducted in-person or by telephone (nϭ36). Email accounted for the rest (nϭ15). Interviews conducted in person or by telephone ranged from 20 minutes to over an hour. All informants were asked the same 8 questions from the interview guide.
Review for commonality. The review for commonalities among responses to the seven questions revealed renewed appreciation for the original research policy as well as general recognition of the need to revise the policy to include current professional priorities. Specific concerns about format, presentation, content, and focus were reflected in the following suggestions for changes: Ⅲ provide a background paper or bibliography to support the statement Ⅲ make it compelling, short; answer the question ''so what?'' Ⅲ point to a research agenda Ⅲ build in accountability Ⅲ emphasize health sciences libraries' research needs Ⅲ be more practical, less theoretical Ⅲ review and update the statement regularly Ⅲ provide role models and examples Ⅲ identify skill sets and domains of library research Ⅲ encourage collaborations Ⅲ recognize small applied research projects as well as grant-funded studies Ⅲ address knowledge management and knowledge translation Ⅲ place the focus on members

Review for recommendations.
The key informant data also revealed seven roles that MLA should take in the new policy statement: leadership, advocacy, collaboration, services, education, publishing, and development of a research agenda. A coda on recognition of success also was noted. Ⅲ Leadership: Interview respondents strongly recom- The Research Imperative mended that MLA lead by engaging the membership in proactively integrating research into their information practice and felt that a research agenda should be tied to the association's strategic plan. In addition to providing an inspired vision to foster a culture of research, MLA should annually fund one or two systematic reviews or member studies to answer the most important library-related research questions on the agenda. The Board of Directors, working closely with the Research Section, should establish mechanisms for follow-up and accountability for action plans included in the policy statement and should encourage sections, committees, and chapters to incorporate research into their own strategic plans. Leveraging MLA's Center of Research and Education (CORE) as a research repository for member-generated research, tools, and datasets was also recommended. Ⅲ Advocacy: MLA was encouraged to promote to members, employers, funding agents, and library educators the importance of integrating evidence into all areas of practice. Interviewees thought the benefits of collaborative research projects and career advancement through the practice of evidence-based librarianship should be stressed to members in addition to recognition through an annual award for the best research paper published in the Journal of the Medical Library Association (JMLA). Further, MLA should identify and feature librarian researcher role models for practitioners at all levels. Recommended advocacy efforts directed to employers included tying research as a value into recruitment efforts, promoting librarian involvement in broader institutional research projects, and profiling best practices in applied research. In reaching out to funding agents (foundations as well as the National Library of Medicine), MLA should stress the need to fund small-scale applied research projects, as well as large-scale institutional or informatics grants.
Recognizing that the research literature of librarianship is weak, MLA should encourage library educators to study the real questions arising in the profession. MLA also should urge library educators to prepare students to rigorously use, create, and share the evidence they will need to be effective, reflective practitioners. Ⅲ Collaboration: Not unexpectedly, key informants thought MLA should work closely with LIS schools to prepare their graduates to build and use a research knowledgebase for the profession. Partnering with library school faculty and information scientists also was encouraged. They thought MLA should explore collaborative studies with other partner associations such as the Association of Academic Health Sciences Libraries (AAHSL), American Medical Informatics Association (AMIA), and SLA on issues of common concern. MLA members should be encouraged to engage in research partnerships with their users and with colleagues in other disciplines. Also, MLA should foster multisite studies of key research agenda questions among its institutional members. Ⅲ Services: Research-related services that interviewees thought MLA could offer included fostering the establishment of a cadre of experienced librarian research-ers who could serve as mentors for members who want to do research. Other services mentioned were a ''translation'' service that would contextualize and inform membership about research findings and their implications; guidelines for evidence-based practitioners and appropriate research methods; recognition for librarian researchers in the Academy of Health Information Professionals; a research methods consulting service; a tool for text mining MLA publications and MEDLIB-L; and support for member access to library literature indexes. Ⅲ Education: Education-related recommendations drawn from informant data referred to actions MLA should take regarding its own educational offerings. The data revealed that MLA continuing education courses should prepare current practitioners to acquire the skills to critically appraise research in their own and their users' published literature as well as prepare members to conduct research. Both using and conducting research require understanding methodology, statistics, analysis, and synthesis. To conduct research, members also might need training in writing grant applications and writing for publication. Further, MLA could help educate members about the environment in which they work by highlighting findings in cuttingedge health research. Ⅲ Publishing: To assure that its research policy statement remains dynamic, MLA should publish periodic progress reports updating the research agenda and recent accomplishments. ''How-to'' toolkits for novice researchers and model position descriptions that include research duties should be developed and shared. Consideration should be given to establishing a ''research only'' publication or section of the JMLA. Ⅲ Development of a research agenda: Various informants suggested that MLA pursue funding or conduct research in areas such as knowledge translation, user information needs, and information retrieval and dissemination. MLA should also establish baseline data on the current state of the profession's research base and measure the outcomes and impact of its actions as a result of the research policy and agenda over time. Some measures mentioned were number and quality of research articles published in the JMLA or presented at the annual meeting and growth, development, and utility of the research portion of CORE. Ⅲ Recognition of success: Interviewees thought MLA would know its research policy statement and the consequent actions it took were successful if all members were ''research literate''; the knowledgebase of the profession was valued; and access to library literature was substantially improved. Thematic consensus. All task force members read the full dataset to discover underlying concepts and relationships. After discussion, the group quickly reached consensus on six themes encompassing critical areas to address in the new policy. Figure 1 outlines the themes described more fully below. Ⅲ Creation of a culture of research: Informants generally expressed the idea that the continued growth and development of the profession and individual profes-

Figure 1
Themes arising from review of key informant data Ⅲ Creation of a culture of research: The continued growth and development of the profession and individual professionals requires the creation and use of a research knowledgebase. Ⅲ Challenges: Conditions (e.g., resources, training, funding) do not currently exist in sufficient quantities to support a research imperative. Ⅲ Domains of research: Critical research questions abound and need to be identified and pursued, and the results shared broadly. Ⅲ Research skills set: Whether published research is used to provide information service to end users or to improve their own practice of librarianship, all health information practitioners need the ability to analyze and appraise the research literature. Ⅲ Roles of stakeholders: Responsibility for the research imperative is shared across the community of practice and includes individual health information professionals, the Medical Library Association (MLA), graduate schools of library and information science, employers, funding agencies, and partners. Ⅲ Measurement of progress: As part of its strategic planning process, MLA should annually identify ways it can advance The Research Imperative and give an accounting of progress in this area.
sionals requires the creation and use of a research knowledgebase. Informants noted, for example, that there is ''a disconnect between research and practice'' (Director 10) and that the research coming out of LIS or informatics programs is too academic or theoretical, ''not in touch with the real research issues . . . in the library setting . . . such as personalized information, human/computer interaction, how people use information, etc.'' (Funding agent 1). Additionally, what passes for library research lacks rigor. Paying more attention to creating and using the health sciences librarianship knowledgebase was a message that came through loud and clear: ''While we are becoming expert at our users' scientific research process, we throw it out the window when we do our own research. There is too much 'how we did it good,' too many case studies, not enough solid evidence'' (Medical informatician 4). Ⅲ Challenges: Informants' comments revealed that conditions (e.g., resources, training, funding) do not currently exist in sufficient quantities to support a research imperative. Funding agents in particular noted that, in the years since the first policy statement was published the number of librarians applying for or included on federal research grants had changed little (Funding agents 1 and 3), and the numbers were still very low. Lack of training, experience, and time were seen as obstacles. Ⅲ Domains of research: Informants, in particular library directors, noted that critical research questions abound and need to be identified and pursued, and the results shared broadly: ''The context of our research today is broadened to include not only the health care field, but also changes in publishing (e.g., open access), society (e.g., bioterrorism), and the practice of librarianship (e.g., informationists)'' (Director 11). Further, ''public health and health policy'' should be the focus of efforts as much as those areas traditionally considered for research-patient care, medical education, and user research (Director 11). Others listed specific questions librarians should be pursuing such as ''development of clinical indicators; curriculum analysis; program evaluation; abstracting and indexing; information storage and retrieval; thesaurus construction; technological innovation; assessment of information needs and information-seeking behavior; and study of the use and impact of information on patients, caregivers, researchers and students'' (Director 6). Ⅲ Research skills set: Whether published research is used to provide information service to end users or to improve their own practice of librarianship, all health information practitioners need the ability to analyze and appraise the research literature. Comments, such as ''We've been so focused on assisting others in getting the evidence to support [their] profession[al practice] that we've lost sight of creating knowledge that would help us in our own'' (Director 10) and ''We have to turn the mirror back on ourselves'' and do with our own literature what we expect health practitioners to do with theirs (Funding agent 1), were common expressions of concern related to research skills and use. The consensus of interviewees was that the research policy needed to speak compellingly to health information practitioners at all levels and in all settings. Ⅲ Roles of stakeholders: Informants felt that responsibility for advancing research is shared across the community of practice and includes individual health information professionals, MLA, graduate LIS schools, employers, funding agencies, and partners. Interviewed hospital librarians, for instance, saw research as a critical survival skill needed to justify to employers everything from continued library funding, to allocation of physical space, to service and resource de- Key informants who had been involved with large research projects and had themselves received grant funding thought the policy needed to stress the importance of collaboration and learning from colleagues. ''Researchers need colleagues with whom they can collaborate'' (Medical informatician 1). Many viewed library educators as logical research partners but noted little had been done to leverage the potential of libraries as research laboratories for information science faculty and students. Many people thought looking to collaborators in other disciplines in one's own institution was most appropriate. One interviewee proposed that library associations should work together to foster both research and collaborations. ''When sponsored or led by MLA, collaborative studies with [the Association of American Medical Colleges] AAMC, AAHSL, research institutes, patient safety institutes, etc. might be more doable'' (Hospital librarian 2). Ⅲ Measurement of progress: The idea that, as part of its annual strategic planning process, MLA should identify ways it can advance the research policy's action plan and give an accounting of progress in this area was a critical theme raised by informants, many of whom expressed frustration at the lack of accountability for actions proposed in the earlier policy state-ment. While they commended the action plan itself, comments such as ''I would like to see some data indicating what has been done since the original publication of the statement. Has it made a difference?'' (Practitioner 4); ''lacks measurable outcomes'' (Practitioner 6); and ''accountability for the action statements needs to be built in'' (Director 10) were frequent.

DISCUSSION
In the more than ten years since MLA's first research policy statement was issued, much in the library profession has changed. The value of evidence-based practice in a broad array of fields, not just medicine, has been widely recognized, as has the importance of collaboration for individual professionals, institutions, and the association. These values were acknowledged in the original policy statement but not fully developed. What the 1995 statement had done so well was highlight the need for librarians to support the practice of evidence-based medicine and suggest ways they could contribute to institutional research, especially in the areas of information policy and information systems. The key informants made it clear that the policy statement needed revision. For example, several of them strongly suggested that the policy needed to encourage librarians to engage in evidence-based library and information practice, not just evidence-based medicine.
Having determined a revision was necessary, the task force decided that the policy should consist of a brief overview or executive summary, which would be distributed widely, backed by a web version with ''foundation'' documents fully exploring each of the six themes. While the summary statement would be distributed widely, the ''foundation'' documents would only be accessible from the association's website. Responding to comments regarding the lack of a bibliography or any evidence of a literature review in the original document, each of the foundation documents was treated as a short research paper with supporting references. An extensive bibliography was included as an appendix. A list of association milestones related to research since publication of the original statement was also included as an appendix because many informants ''found it a bit frustrating . . . to look at all the specific recommendations aimed at MLA and not to know for sure which have or have not been accomplished, and what the impact was of those that were'' (Director 7).
Several informants also mentioned the need to identify role models and examples of successful research initiatives involving librarians. To address this, the task force invited seven librarians to share their experiences with both large scale and applied local research projects. These interviews or vignettes were videotaped and were interspersed among the foundation documents at appropriate places.
The task force also decided that all recommendations should be couched in terms of MLA actions, not those of other groups or organizations. While the importance of stakeholders was recognized, MLA's role in relation to them should be the focus. Further, the task force chose not to integrate the recommendations into an action plan as the first statement had done, but to make the action plan a separate appendix with short-, mid-, and long-term objectives. The task force hoped that this would keep the basic document relevant over time and allow the action plan to be updated and revised annually.
The final document was presented to the Board of Directors in May 2007 and received final approval in September 2007.

STUDY LIMITATIONS
Study limitations included response and analysis issues: first, the 2 open forums engaged a small number of MLA members, and the low response to the followup MLA-FOCUS items could be interpreted as lack of interest rather than approval. Second, the methods of interviewing and transcription of the key informant interviews varied among the various task force members. Informants who were interviewed face-to-face or by telephone tended to give extended answers to the interview questions, exploring implications and offering detailed suggestions for improvement. Informants who responded by email, 30% of the total, generally gave much shorter responses and made many fewer suggestions. Finally, because transcriptions were completed by each interviewer and not a single individual, variability in quality and completeness of transcripts was inevitable.

CONCLUSION
The revised policy statement, The Research Imperative, emphasizes the need for all health information professionals to incorporate evidence into their information practice as a survival skill. It envisions a future in which it is routine for librarians to provide essential research support, including critical appraisal, for health practitioners; in which institutional support for information resources and services is strong because of data collected through evaluation studies that clearly demonstrate their value to and impact on users; and in which the health information knowledgebase is a rich source of evidence for information practitioners.
The task force saw The Research Imperative as MLA's commitment to creating a supportive culture in which members can grow and develop as users of research and increasingly as creators of the evidence on which the profession itself depends. This commitment includes developing a research agenda focused on the domains of health sciences librarianship and exploring questions dealing with effectively identifying, contextualizing, synthesizing, sifting, structuring, disseminating, and using information to improve health care; providing training to gain the necessary skill set; and creating incentives for members to make the leap to evidence-based practice.
In addition to providing a vision, rationale, and guidance for creating a culture of research among health information professionals, The Research Impera-tive is a statement of values and, as such, is an expression of the highest aspirations for the profession, for individual librarians, and for the association.