Abstract

Objective – To determine the perceptions and information behavior of institutional repository (IR) end-users.

Design – Semi-structured interviews.

Setting – The interviews were conducted over the telephone.

Subjects – Twenty end-users of five different IRs were interviewed for the study. Seventeen of the interviewees were recruited via recruitment forms the researchers placed on IR homepages and the other three interviewees were referred to researchers by IR managers.

The interviewees’ academic backgrounds varied, including six undergraduates, four masters’ students, three doctorial students, five faculty, and two library or museum staff members. They represented disciplines in Arts and Humanities (5), Science and Health Sciences (10), and Social Sciences (5). Fifteen of the 20 interviewees were recruited through their own institution’s IR. All except two of the interviewees had used the IR for which they were recruited less than six times.

Methods – Forty-three potential interviewees were recruited using web recruitment forms and IR manager recommendations. Researchers subsequently excluded 23 (53.5%) of the interviewees because they were primarily IR contributors rather than end-users, or could not be reached by phone.

Twenty interviews ranging from 17 to 60 minutes were conducted between January and June 2008. The average interview time was 34 minutes. The recordings were transcribed then analyzed using qualitative data analysis software NVivo7. Coding categories were developed using both the original research questions and emerging themes from the actual transcripts. The final coding scheme had a Holsi Coefficient of Reliability of 0.732 for inter-coder reliability.

Main Results – Researchers identified six common themes from the results:

How do end-users characterize IRs?

While most interviewees recognized that there is a relationship between the IR and its host institution, their understandings of the function and content of IRs varied widely. Interviewees likened the IRs they used to a varying array of information resources and tools, including databases, interface, server, online forums, and “static Wikipedia” (p. 27). Furthermore, six of the interviewees had never heard of the actual term “Institutional Repository” (p. 27).

How do end-users access and use IRs?

The most common methods of accessing IRs included selecting the link on their institution library’s website and Google searches. Many interviewees found out about the IRs they are using through recommendations from professors, peers, or library workshops. Other interviewees found out about particular IRs “simply because a Google search had landed them there” (p. 29).

Interviewees’ preferred method of interacting with an IR were divided between browsing and keyword searching. However, these preferences may have been the result of an IR’s content or interface limitations. For instance, some interviewees expressed difficulties with browsing a particular IR, while another interviewee preferred browsing because “there wasn’t much going on” when searching for a specific topic of interest (p. 30).

For what purposes do end-users use IRs?

Interviewees commonly cited keeping abreast with research projects from their own university as a reason to access their institutions’ IRs. Student interviewees also used IRs to find examples of theses and dissertations they would be expected to complete. Identifying people doing similar work across different departments in the same institution for collaboration and networking opportunities was another unique purpose for using IRs.

How do end-users perceive the credibility of information from IRs?

Many interviewees perceived IRs to be more “trustworthy” than Google Scholar (p. 33). In their view, an IR’s credibility was assured by the reputation of its affiliated institution. On the other hand, many interviewees viewed a lack of comprehensiveness in content negatively when judging the credibility of an information source, which placed most IRs in a less favorable light.

Additionally, researchers noted conflicting assumptions made by interviewees about IRs in the evaluation process for their content. Some interviewees believed all the content of an IR has been vetted through an approval process, while others distrusted all IR content that was not peer-reviewed.

To what extent are end-users willing to return to an IR or recommend it to their peers?

The great majority of interviews indicated they were likely to use IRs again in the future, and nearly all indicated they would recommend IRs to their peers. However, most interviewees did not know of any people using IRs. The few interviewees who did often knew of IR contributors rather than end-users.

How do IRs fit into end-users’ information seeking behavior?

Many interviewees noted that IRs provided them with content that was not commonly available through traditional publishing channels, including conference papers and dissertations. Others felt IRs made content available more quickly than other information sources. However, the results also suggested that most interviewees did not include IRs in their routine research process.

Conclusion – This study identified current end-users’ perceptions of IRs and highlighted several areas for future IR development. Areas of improvement for IRs included intensifying publicity efforts; increasing content recruitment; making content recruitment policies more transparent; and improving appearance and navigation functionalities. The findings also suggested new directions for IR marketing, such as emphasizing on the networking and collaborating benefits of using IR.

Commentary

This exploratory study uncovered several insights for IR development. Study results indicated end-users were largely unfamiliar with the purpose and scope of IRs. A significant portion of end-users surveyed were also unsatisfied with the collection size and usability of IRs they have accessed. These findings provided valuable directions for IR improvement, especially in user-experience related areas such as interface design and marketing. Nonetheless, this study was exploratory and its findings were meant to generate new research ideas and encourage further scholarship, not to serve as generalized conclusions. There were also several shortcomings in this study that future research could improve upon.

One flaw of the study lies in its subject recruitment through sign-up forms posted on IR homepages. As the authors themselves noted, past studies found that majority of end-user reaches IRs via Google or Google Scholar, which bypassed IR homepages. Since majority of users from the five IRs were excluded from the recruitment process, then, one cannot conclude the interviewees’ comments were representative of perceptions of general IR end-users.

Moreover, while the researchers noted difficulties with differentiating between IR end-users and contributors, their actual methods for distinguishing them were not specified. Five (25%) of the interviewees were both IR end-users and contributors, and this inclusion could have negatively affected study results. For instance, part of the investigation included interviewees’ perception an IR’s content quality, and interviewees with contribution experience and familiarity with an IR’s content recruitment policy would likely have had a different perspective than end-users.

Lastly, the researchers’ rationale for selecting the five particular IRs for recruitment was not specified, nor did the researchers identify these IRs. Providing access to the IRs reviewed in the interviews would allow audiences to better understand some of the interviewees’ comments. In one instance, the researchers noted conflicting interviewee opinions on whether IRs were better for browsing or searching. Such preference variations could have been influenced by specific IR designs that interviewees were familiar with, but this hypothesis could be not verified since the IRs discussed were not identified.

Due to these limitations in data collection, the overall validity of this study is less than 75% based on the EBL Critical Appraisal Checklist (Glynn, 2006). This validity score suggests readers should not use the results for generalized conclusions. Even so, this study provided valuable contribution to current literature because it highlighted unique challenges face by IR end-users and provided directions for future IR designs.

References

Glynn, L. (2006). A critical appraisal tool for library and information research. Library Hi Tech, 24(3), 387-399.