3. Universities and EDTs

The advantages of ETD systems for universities

      For universities, an ETD program has numerous advantages, in addition to the grand one of helping build a worldwide collection of millions of graduate research reports. First, it is a way for the research carried out in connection with their graduate programs to become visible to large numbers of interested parties around the world. High quality ETDs may add not only to the reputation of the students preparing them, but also to the faculty, research groups, laboratories, centers, departments, colleges, and universities involved. Even on a single campus, other students engaged in research, as well as instructors seeking interesting examples for classes, may benefit from each ETD added to the local collection.

      Second, an ETD program may save time, labor, and funds that would be devoted to more conventional processing of paper TDs. If a campus switches from paper to electronic submission, there are savings in library shelf space, binding, shelving, hauling and shipping, and reductions in the costs associated with checking and cataloguing. Based on experience at Virginia Tech, the value exceeds $10,000 per year.

      Third, an ETD program helps lead to improvements at universities regarding digital library infrastructure. Though the number of works and accesses are only moderate relative to larger digital libraries and online collections, a full implementation of an ETD initiative constitutes a complete digital library application. Indeed, the planning, training, implementation, and operation of an ETD program can be thought of as a complete digital library case study (Fox 1999b). It should be easy afterward to undertake other digital library projects. Conversely, if a campus has digital library efforts underway, adding ETD services should be a relatively easy enhancement.

      Fourth, and most importantly, ETD programs may raise the understanding on a campus of key concepts. There may be increased awareness of the value of multimedia methods to express research results. There may be more understanding of digital libraries, more support for digital preservation programs, more willingness for authors to submit their works into open archives, and more emphasis on the development of skills related to searching, accessing, and re-using knowledge resources. There may be increased discussion and understanding of issues related to intellectual property rights, publishers, the value of university research, and the various ways in which research results can be disseminated. There may be increased valuation of information literacy, and expanded support for graduate programs.

Develop an ETD program in your university

      Campuses interested in ETD programs engage therein when there is sufficient leadership and initiative. If a concerted effort is made, the entire process may be completed in less than half a year, though some campuses may gradually shift toward ETDs over several years.Typically, an ETD program must be developed as a team effort involving those involved in graduate education, library and archive operations, and computing / information technology support. The relative roles of these three groups, and others involved as per campus situations, depend on local policies, procedures, resources, skills, and initiative. While a particular campus can learn from the experience of active institutions in NDLTD, or work in concert with neighbouring or peer institutions as part of co-operative programs, local action is nevertheless needed for this effort that deals with student education and campus infrastructure.

      Some universities have a strong graduate program, in some cases run from a graduate school or as part of a division of research and graduate studies. Others have a commission responsible for graduate activities, or control such efforts through a faculty senate or other governance group. In some cases, separate discipline or profession oriented schools or colleges (e.g., a College of Engineering or a Law School) control graduate efforts and manage all activities related to TDs. Accordingly, decisions to engage in ETD programs may be decentralized, and a part of a campus may support ETDs before other groups, or a representative group may deliberate regarding any campus-wide projects. In any case, from the graduate program area the key contributions are to expand graduate education to support the initiative, and to change policies to allow ETDs in addition to, and eventually instead of, paper TDs.

      Libraries often are the active party in launching an ETD initiative since they usually receive TDs, catalog them, and make them accessible to local readers or to others through interlibrary loan services. Many libraries also assist students in learning to use digital libraries. They may provide archival services, or there may be a separate campus archive – in any case digital preservation is often of concern.

      Computing or information technology groups may run digital library systems, or may support such efforts in the library. Through offsite storage and backup services they may help manage digital preservation activities.

      Any of the three groups may run education or training programs so that students understand the local ETD program and develop skills for creating and submitting ETDs. Special support for multimedia is most often provided through computing or information technology groups, though that may be through a special media centre or in the library. Control of the overall process often is in the hands of the graduate program, though it may be managed in the library.

      By way of example it may be of interest to consider the situation at Virginia Tech. The Graduate School runs the program, setting policies. The Computing Center hosts some of the computers and Web sites involved, though most are in the Library. The New Media Center runs training workshops and supports walk-in students needing help. Students upload their works to a Library computer, running locally developed workflow and database management software (freely available for other campuses to adapt), which allows access by both Graduate School and Library personnel who review and approve submissions for subsequent cataloguing. The accessible digital library is run by the Library, which also assumes responsibility for long-term preservation, collecting a $20 archiving fee for this purpose. In the case of doctoral dissertations, UMI is paid with student funds for works to be uploaded into the UMI collection as well. Though there have been minor shifts in responsibility since the time this workflow was put in place in 1996, the whole operation proceeds smoothly, and regular surveys not only support tuning but also show general satisfaction with the program.

The key concerns for universities and their resolutions

      Since an ETD program calls for change, there are inevitable complaints and concerns that arise. However, based on the experiences of NDLTD members, there are reasonable solutions for all problems raised (Fox, Eaton, McMillan, Kipp, Mather, McGonigle, Schweiker, & DeVane 1997a).

      First, there are concerns regarding ownership of intellectual property rights related to ETDs. In most institutions, ownership of rights for an ETD rests with the author. However, in some institutions, the institution itself may claim or request assignment of such rights. When research results reported in an ETD arise through funding by a particular sponsor, conditions agreed to when that funding was accepted may have an effect on the rights related to the ETD. Eventually, though, it will be clear what party or parties own the rights on the ETD, and it will be known if there are any special constraints that must be met. In addition, it should be known who are the stakeholders who will advise about rights management issues, for example, legal counsel, intellectual property rights offices, faculty supervising the research, or colleagues involved in related research.

      Second, there is the matter of what access is allowed to an ETD. Such a decision is of concern to the abovementioned stakeholders. They may decide differently for any part of an ETD, since digital library technology can allow separate access controls to be in effect as appropriate for different portions (e.g., a chapter that covers information that appeared earlier in a journal, a chapter submitted for possible appearance in another journal, an image provided for scholarly study and criticism by a third party, or a literature review that discloses no new methods but instead is likely to be of interest to the general public). One decision, promoting scholarly communication, is to make content freely available. Another decision, satisfying desires to limit access to the local campus, may be to restrict access to the university community and its library patrons. Strictest control, such as when patent protection is sought, is to avoid disclosure except to those supervising or reviewing the ETD as required for approval. Note, however, that in the interest of facilitating access, at least in the long term, any of the schemes for control may have a time limit, though possibly allowing renewal.

      Third, there is the question of how ETDs relate to publishers. For most students, there are no publications involved, so this is a non-issue. For students in the humanities or social sciences, for example, where advancement often hinges upon publishing a book, usually involving a limited print run, discussion with prospective publishers should proceed prior to deciding about ETD access. Available data suggests that it is very rare for a student to publish a book that is at all similar to their TD, and that there is little evidence that public access to an ETD will hurt future sales of an eventual published book that relates. Nevertheless, students working on a book may decide to limit access to the university community for a reasonable period if so advised by a publisher. On the other hand, when a student works in other fields, such as the hard sciences, they may consult with the publisher of a journal to determine if there is a problem regarding making their ETD publicly available. If their ETD has similar content to an already published article, they should secure permission from the copyright holder for the article, and typically will add an acknowledgement. If they hope that their ETD will lead in the future to a journal article, they may find that publishers have no concern with the ETD being available, or else may be required (for a time) to limit access, typically to the university community. Eventually it is hoped that as NDLTD expands, and ETD programs become better understood, then all publishers (not just those on a list that have notified NDLTD) will see how different the genre are, and will allow free access to ETDs.

      Fourth, there is the issue of plagiarism. It is true that if ETDs are readily available then people may copy from them and claim others’ works as part of their own. However, search technology makes it possible to detect such copying (even more so than is possible today, where so many theses available only on paper remain unknown to most scholars). Further, TDs are supervised by groups of faculty, who should be knowledgeable about their students’ research, and who often carry the authority of honor codes and other strict rules. Thus, students who commit plagiarism may run a terrible risk of detection and severe punishment.

      Fifth, there is the matter of cost. Running an ETD program involves personnel to propose, publicize, initiate, refine, and institutionalize the activities. If lessons are learned from those already engaged in successful ETD activities, startup costs can be reduced, and smooth operation can soon occur. If a campus is committed to having knowledgeable graduate students able to prepare electronic documents, who are well prepared to be scholars in the electronic age, there is little extra load needed for implementing an ETD program. Indeed, as was mentioned in Section 3.1, when ETDs instead of paper TDs are required, there should be net savings relative to old processing methods. However, if a paper form is managed in addition to an ETD, or if ETD preparation is by university staff instead of by students, there will be small additional work incurred. Typically, any extra work can be carried by existing staff in connection with their normal duties, and certainly involves no more than the effort of a part-time employee.

Evaluation of EDTs at the university level

      Implementing an ETD program should be accompanied by formative evaluation efforts to ensure that needed improvements and refinements are made as soon as possible. At Virginia Tech, data is collected whenever feasible at workshops, when ETDs are submitted, when people wish to access the ETD collection, and periodically from students after varying lengths of time following graduation. No student has yet reported a problem with a publisher resulting from their submitting an ETD.

      Generally, quantitative and qualitative results have been quite positive. Most ETDs are accessed hundreds or thousands of times as opposed to the normal case of TDs that are accessed much less than ten times per year. Most students are in favor of the program. Some have made new contacts or been pleased that their works have been of interest to or impressed others favorably. Workshops (usually for beginners, though sometimes for those interested in advanced topics) are generally found to be helpful. A very small number of students, typically those with little facility in electronic publishing, are unhappy with the initiative. They argue that they should not be required to submit an ETD, and complain about extra work involved. It is likely, however, that they would oppose any effort making computer and information literacy mandatory.

University policy initiatives

      University ETD programs must fit into the general schemes of local, regional, and national initiatives for education and scholarly communication. Many of those, such as the NCSTRL project for computing to provide access to technical reports (Davis & Lagoze 2000; Lagoze 1999; Leiner 1998), function as federations supported by distributed processing technology. NDLTD similarly assumes that the overall collection is composed of a number of repositories that can be harvested from, or can participate in a federated search service (Powell & Fox 1998). The organizing principle behind each repository may vary as needed.

      Most NDLTD members are individual universities that have elected to join and participate as an institution. Some begin that process by way of a pilot effort in a particular campus sub-unit that is ready to support student submissions before campus-wide infrastructure and policies are in place. On the other hand, some groups of universities join together, building upon related initiatives or practices for collaboration, to develop ETD programs as shared efforts. For example, OhioLINK supports ETD efforts for all interested institutions in the state of Ohio. In Catalunya, a consortium involving universities and libraries agreed to manage the regional and language-related group of interested institutions.

      University de Lyon in France and University of Montreal in Canada are cooperating in a Francophone effort to encourage ETD activities in the French-speaking world. This is analogous to efforts involving ISTEC and OAS (described in this monograph) to support efforts in Latin America and Ibero-America. In all these cases, special support by interested organizations, in most cases involving small amounts of funding for programs, has facilitated workshops and training. However, the vast majority of the costs of shifting to ETD programs are carried by individual universities and their staff involved in that work.

      At the national level, small amounts of funding have supported launching ETD activities. As was discussed in the previous section, regional support by SURA and national funding by the Department of Education led to the initial spread of the concept in the Southeast and then to the rest of the USA. Funding also has supported national programs in Germany, Australia, India, and most recently, through the Mellon Foundation, in South Africa. Generally, such funding is limited in duration since mature programs are self-sustaining.

      While almost all NDLTD-related universities allow free access to works, at MIT a different financial arrangement is involved. Some students prepare ETDs, while others still submit paper TDs that are scanned, yielding PDF files containing page images. Access to the metadata for the entire MIT collection is free, as is display of PDF files on screen, but MIT collects payment through an e-commerce scheme for printing of TDs from its repository. Requests for old TDs not in the electronic collection lead to scanning of those works so they are added to the collection, resulting in a partial retrospective conversion of in-demand work. Of course commercial organizations like UMI, Diplomica.com, Dissertation.com, and others also must have business plans to allow them to provide services to students and universities related to TDs. It must be remembered, however, that the essence of NDLTD is to support education of students, sharing of research results, building university infrastructure, and other causes that only relate indirectly to whatever and however many other access schemes arise with regard to the ETDs that students learn how to produce.