Toward successful institutional repositories: Listening to IR staff's experiences



This poster presents a study of what is required to plan, implement, maintain, and sustain successful institutional repositories (IRs). This study is part of the MIRACLE (Making Institutional Repositories a Collaborative Learning Environment) Project, which aims to identify the factors contributing to the success of IRs and effective ways of accessing and using repositories. The data for this study were collected through telephone interviews with IR managers and staff members. Interviews focused on participants' motivations, experiences, and viewpoints in regard to IRs, among other IR deployment issues. The transcriptions of interviews were analyzed using content analysis. The findings indicate that IRs are more than the sum of their component parts (i.e., collections, systems, policies, staff, contributors, and end-users); rather, they are a knowledge sharing and content management tool that works together with other types of open access infrastructure in colleges and universities.


This poster presents the latest findings of the MIRACLE (Making Institutional Repositories A Collaborative Learning Environment) Project. The MIRACLE Project is an IMLS-funded three-year project whose goal is to collect the experiences and viewpoints of various IR stakeholders (from the library directors and IR staff members to contributors and end-users who are depositing, searching, and using IR content) and to then identify specific factors that contribute to the overall success of IRs. Project investigators seek to identify models and best practices in the administration, technical infrastructure, and access to repository collections.

The MIRACLE Project consists of five distinct phases of data collection. During the first phase, the authors conducted a nationwide census of 4-year colleges and universities in order to find out about their involvement with IRs. More than 2,000 Library Directors and other staff were invited to participate in this census. The results based on 446 respondents have been reported in the authors' earlier publications (Markey, Rieh, St. Jean, Kim, & Yakel, 2007; Markey, St. Jean, Rieh, Yakel, & Kim, 2008; Rieh, Markey, St. Jean, Yakel, & Kim, 2007; Yakel, Rieh, St. Jean, Markey, & Kim, in press). This poster focuses on the second phase of the MIRACLE Project which consisted of follow-up interviews with 36 of the census respondents. The MIRACLE Project also encompasses interviews with IR end-users about their search and use experiences, an experimental study of end-user searching of IRs, and a series of case studies at a handful of model IRs.

Research Methods

Four semi-structured interview protocols were developed, one for each stage of IR development (implementation, planning and pilot testing, planning only, and no planning for an IR). Project investigators selected 36 of the 176 census respondents who had indicated on their census form that they would be willing to participate in a follow-up telephone interview. We used purposive sampling in order to select these 36 interviewees, using criteria such as the potential interviewee's job title, the stage of development of the IR, the extent and types of content housed in the IR, and the size, Carnegie Classification (The Carnegie Foundation for the Advancement of Teaching, 2006), and geographic location of the parent institution. The interviewees included 11 library staff members, 9 library directors, 4 assistant/associate library directors, 4 archivists/directors of archives, 4 heads/directors of library units, 3 ClOs, and 1 associate dean for research. Our interviewees were also distributed across various stages of IR development: 14 had already implemented an IR, 10 had completed some planning and pilot testing, 7 had done planning only, and 5 had done no planning for an IR.

All interview recordings were transcribed and the transcripts were imported into NVivo 7, qualitative data analysis software. The content analysis yielded a number of issues and concerns discussed by the interviewees, resulting in 24 major categories and 37 subcategories. Some of the major topics that emerged from the interviews included impetus, objectives, and mission for the IR; perceptions of IRs; content and content recruitment; policies and access; preservation; end-users and uses; IR services; intellectual property rights; evaluation and metrics; and sustainability.

Preliminary Findings

Content analysis of the interview transcripts is revealing a number of interesting findings. First, we have found that the IR is much more than a single repository. It is indeed more than the sum of its component parts, including the collections it houses, the system upon which it is built, the policies that govern it, the managers and staff that build and maintain it, the contributors that both benefit and are benefited by it, and the end-users it serves. The IR staff members who participated in our interviews view an IR not as a stand-alone system, but rather as a knowledge sharing and content management tool which works together with other types of open access applications and infrastructure. Most of them agree that it is inevitable for academic institutions to deploy IRs in order to provide access to and to preserve digital materials; however, they are also aware that the success of IRs requires long-term institutional commitment to visibility, accessibility, and availability of the IR.

Many participants focused on capturing and preserving digital information produced in their institution while paying less attention to developing a service model for their community. They often mentioned simply accessibility to full-text as an important aspect of the service they offer. However, most could not comment much on “user support” as their effort is concentrated on providing the “vehicle” to publish scholarly papers. Only one institution has the ability for their members to use RSS feeds through which they can receive emails about a particular type of content that they are interested in. Another IR staff member mentioned that she is talking with “people on campus” asking what the IR can do for them within the context of their own discipline. She has noticed that people provide very different answers from each discipline about the best IR service model for them, thus her institution has chosen to focus on making widely available and preserving those materials that are outside the traditional publishing path.

When the participants selected software for their IRs, they often did literature searching, talked with other libraries, and pilot-tested about three IR software systems that are widely adopted and then chose one, taking various criteria into consideration beyond technical features, such as ease of system migration, maintenance, technical support, and standards. Smaller academic institutions, with limited staff, tend to purchase commercial software for their IR rather than open source software which requires more knowledge of programming and networking expertise. IR managers and staff who chose to develop their own IR system did so because they believed that they could explain to people how it works and how to use it rather quickly and effectively.

DSpace was selected most widely among participating IRs because “open source was very appealing to us,” it is “well developed”, it works “very fast,” and it works out “out of the box.”

Policy development for IRs appears to be still in its infancy. Many participants responded that they have drafted IR policies, but that they plan to revise them later. They tend to use policies as a framework that, once in place, can be used to have “a serious conversation with anybody” on campus. One participant explained that they first drafted IR policies based on the hypothesis of “how things might work” and have since revised them four times already during their planning stage as they became able to add more detailed and more specific policies in terms of what they would or wouldn't do with different file formats in terms of preservation. Overall, there is consensus among participants that policies are essential in determining who can access, who can read, and who can write to specific areas of the IR. They consistently pointed out that policies need to reflect the unique needs and environment within their organization. Participants expressed concern about intellectual property rights; however, they tend to be lenient toward these issues, showing some “wait-and-see” attitudes.

Data analysis is still underway. We are particularly interested in discovering the perceptions of IR managers and staff as to the potential benefits of IRs, the factors that can facilitate or impede the success of an IR, the weaknesses and limitations that can stifle the potential growth of an IR, and the various factors that can either promote or threaten the sustainability of an IR.


Support for the MIRACLE Project is provided by the Institute of Museum and Library Services (IMLS) through its National Leadership Grant Program (LG-06-05-0126-05).