Request for Widespread Access to Electronic Theses and Dissertations

 Edward A. Fox and Gail McMillan --- October 1997

1. Theses and Dissertations - An Important Genre

Students engaged in research at the bachelor, master, and doctoral levels have prepared theses or dissertations (TDs) for hundreds of years. These works are usually written in partial fulfillment of degree requirements, and contain valuable information, including focused literature reviews and details on research not generally made available elsewhere. This genre supplements other types of documents, like reports, conference papers, journal articles, textbooks, and monographs. For example, ACM has recognized the importance of dissertations by giving a prestigious annual award for the best dissertation in the computing field.

2. Numbers and Distribution of Theses and Dissertations

Access to dissertations has increased in recent decades as a result of services provided by UMI. Their database contains about 1.5 million abstracts. Their microform archive allows printing on demand of microform or paper copies. While this service is useful, the high cost of copies and other concerns lead to relatively few purchases, almost never more than 7 per year for any one title.

UMI receives very few bachelors or masters theses, however, other than theses from Canada. A recent study by OCLC indicates that libraries have catalog records for about 1.5 million theses, and Council of Graduate Schools data indicates that each year, in the U.S. alone, there are over 350,000 masters degrees awarded. Except through personal efforts, or interlibrary loan, the thousands of theses prepared each year are rarely distributed. Circulation data from the Virginia Tech library indicates that the average number of times a thesis or dissertation is checked out per year is small (2 or 3, respectively, during the first 6 years after completion).

In 1987, UMI began discussions with Virginia Tech and others about moving toward electronic theses and dissertations. While a number of universities have explored this idea to some degree in the intervening years, only recently has technology been available to support a federated effort in this direction.

3. NDLTD

With funding for 1996-97 by the Southeastern Universities Research Association (SURA) and for 1997-99 by the U.S. Department of Education, Virginia Tech has led an initiative to develop a Networked Digital Library of Theses and Dissertations (NDLTD). This is aimed at improving graduate education by ensuring that students learn about publication issues (including how to prepare their own electronic theses or dissertations - ETDs) and about using digital libraries. At the same time, the ETD initiative is helping develop an important digital library that ultimately will include millions of full-text or even hypermedia documents.

In fall 1997 there are over 20 universities (in Africa, Australia, Europe, and North America) involved in NDLTD and scores of others investigating this initiative. As interest spreads, this could lead to a dramatic increase in knowledge sharing, and have a lasting impact on future generations of scholars. In addition, if all new scholars learn about electronic publication issues and use of digital libraries, this should reduce operational costs of publishers.

Work toward the NDLTD has broad support from a variety of organizations. Substantial donations have been provided by IBM, Microsoft, and Adobe. A large Steering Committee advises on policy issues. Members include the sponsors and supporters mentioned above as well as NSF, World Bank, National Library of Canada, SOLINET, CIC (Big 10), and the University Thesis Online Group in the United Kingdom.

4. Requesting Assistance of Publishers

As lead institution in the NDLTD, Virginia Tech requires all students to submit their theses and dissertations in electronic form, starting January 1997. Publicity about this has led to coverage by the Chronicle of Higher Education, National Public Radio, the NY Times, and many regional newspapers. Unfortunately, media coverage has tended to focus on controversy, and to picture the matter as one of conflict between universities and publishers. That is not the intent of the NDLTD, which focuses on graduate education and further developing a valuable resource that complements other forms of publication.

Theses and dissertations are very different from other forms of publication. They are much larger and more detailed, and as electronic works can easily include multimedia files, databases, simulations, and other components that may consume large amounts of space. While reviewed by a group of local advisers, they are not refined and certified though a peer review process. These characteristics, especially their very size, makes it unlikely that free access to ETDs would hurt publishers of other types of documents. Even when part of an ETD is similar to a conference paper or journal article, it seems unlikely that readers would prefer digging through large ETDs to find information when they can read a more carefully and tersely framed journal description. Accordingly, we believe that the NDLTD should be allowed to provide free access to ETDs. We request that publishers reflect carefully on this situation, and assist universities to carry out their educational missions by supporting this policy.

5. Approval Form and Publisher Policies

To handle special situations and to protect students and faculty until publisher statements can be made public, Virginia Tech has asked students and their committees to fill in an Approval Form when their ETD is complete (see http://etd.vt.edu/submit/). This document indicates that the student has obtained publisher permission if any third party copyrighted material is to be included and made accessible. It also lets the student and committee choose if access to their work (or parts thereof) should be worldwide (which we encourage), or restricted to the campus (in case a publisher has first publication or other policies that make this necessary). Access restrictions may even govern ETD sharing through interlibrary cooperation and through services offered by UMI. Finally, the form indicates when the university should ask if access can be broadened, e.g., after 1 year or after 3 years.

The Approval Form allows publisher policies to be implemented through actions of students and their committees and use of digital library access controls. However, many students do not understand publisher policies, and so choose to limit access to their campus out of fear that they will not be able to publish related works in journals or other forums if their ETD is widely accessible.

It would be infeasible for every student to contact individually all editors that might possibly deal with a journal submission, and ask them to explain their policies regarding ETDs. Accordingly, we hope to have publishers make statements on this matter that can be widely publicized. Some possible policies of publishers might include:

1. allowing ETDs to be freely accessible, continuing the practice that has been in place with TDs;
2. allowing ETDs to be freely accessible, when not very closely reated to a work submitted to a publisher;
3. allowing most of an ETD to be freely accessible, and the rest (e.g., a chapter) which closely relates to a work submitted to a publisher to be made freely accessible also one year after the publication has appeared; and
4. changing copyright release forms so that theses and dissertations can be made freely accessible (i.e., so that students retain rights of this type for their ETDs and universities fulfill their mission to share knowledge).

The Publications Board of ACM, the First Society in Computing, has adopted a policy according to option 1 above. It would be helpful if other organizations follow suit. We hope at the very least that all publishers will support higher education by allowing free access to all ETDs, if necessary, after a delay of a few years following their release of related articles or books.

As NDLTD grows, other institutions will evolve similar approval forms and policies to those of Virginia Tech. Statements of publishers, like 1-4 above, will make it clear to students, faculty, and universities how to decide about providing access to ETDs. Please help NDLTD in particular and graduate students in general through a clear and public support statement encouraging access to ETDs!