2. A survey of ETD preservation at 16 institutions

2.1 Purpose and method

In September and October of 2002 a survey on the preservation of ETDs was carried out.^<1> A list of institutions with either the option or the requirement to submit an ETD in fulfillment of graduation requirements was gathered. Policies of the following institutions were examined:

• American universities in the Digital Library Federation
• North American universities listed as official members of the Networked Digital Library of Theses and Dissertations (NDLTD).
• Association of Research Libraries (ARL) member institutions ranked above NCSU Libraries (i.e., 31 and above in the 2001 ranking).

We learned of additional institutions through an inquiry on the ETD-L listserv. Our final list included 17 institutions (NCSU is not included in this list or in the results below).

Through Web research and phone calls, individuals familiar with ETD preservation practices and workflow were identified at each institution and asked to participate. We offered to share aggregate results of the survey with participants. The survey was administered by phone or e-mail, depending on each participant's preference, in September 2002.

We received responses from every targeted institution except one. The 16 survey participants are listed in Appendix A. The questions asked in the survey are attached as Appendix B.

2.2 Results

Results are presented in the following topical areas: general ETD policy, electronic file format, electronic file backup, paper copies, physical copies of dissertations, differences in processing master's theses and dissertations, migration plans, and workflow and coordination. Because of space limitations, some results are summarized without further discussion.

As agreed with survey participants, responses are reported without attribution to particular individuals or institutions.

2.2.1 General ETD policy

Dissertations in electronic form: Required at 11 institutions; optional at 5 institutions.

Master's theses in electronic form: Required at 7 institutions; optional at 6 institutions; not an option at 2 institutions.^<2>

2.2.2 Electronic file format

9 of 16 institutions require PDF for the main document.

7 of 16 accept PDF or other formats (mainly HTML) for the main document.

- 3 institutions accepting HTML either have strict guidelines for it or convert to PDF.

- Nonstandard formats accepted include Postscript (1 institution) and XML or XHTML (1 institution)

2.2.2.1 File format of the main document

Seven respondents expressed reservations about PDF. At the same time some respondents recognized the value of PDF and the barriers to implementing an XML-based ETD program. Their responses follow:

"[Some on our ETD committee] want to move to something that's more flexible, something that can be migrated more easily than with PDF... Acrobat 5 is supposed to be a little more flexible where you can save as HTML or XML. I'm not terribly excited about it."

„We're hoping [the use of PDF] is a temporary situation... I don't want to be beholden to Adobe."

"As soon as export to XML export works we'll try to do that... We assume that PDF will be going away and that we will be converting to XML sometime but not at the present time. We're committed as much as possible to open source standards."

"...There don't seem to be too many easy answers. PDF is easy for the student to produce and to give to the library but it's difficult to believe that the library could sustain that for any amount of time. On the other hand creating HTML or XHTML puts an enormous burden on the student and they would likely need support from the library that we at least now don't feel that we could provide. Although the end result would be more sustainable it requires a whole magnitude of effort from the student and therefore from the library as well to support it. So being able to influence which format is best is a difficult question to answer."

"We're looking forward to SGML or XML as being a format for both preservation and access but we're not going to ask our students to put it into XML until there's an editor [i.e., software to simplify the process of saving in XML format]."

"...In terms of people doing full-blown XML, the time has not been ripe, particularly at the student level. [The current system is] already like pulling teeth with some students. There's a learning curve on producing XML documents. [XML features are] just coming out on WordPerfect and MS Word. But it's not an easy thing. We know that this holds a lot of promise for the future - there are all the goodies that go along with XML including migration. If XML becomes the technology of the future, we should be able migrate our collection. This is a leap of faith, but we took a leap in 1998 [by starting to accept ETDs].“

"We realize that this format [PDF] is not going to last forever."

2.2.2.2 Supplementary files

Four respondents expressed concerns about supplementary files. One said that if authors are „attaching multimedia and getting fancy with their supplementary files they ought to do it at their own risk.“ Another said that „The document should work without ancillary materials... [These are] just an enhancement.“ If supplementary files fail, it was explained, the main document can still stand alone as a complete and coherent thesis or dissertation. The ETD website of a third institution states: „The advantages of access must be weighed against the value of long-term preservation of textual content. The candidate and supervisor may wish to consider limiting multimedia to appendices that enhance but are not required for comprehension of the thesis“ (emphasis added). The fourth institution simply does not accept files in any format other than PDF: „We don‘t like the multi-files, trying to sort them out. We‘re trying to go with a very high-quality PDF that is appropriately backed up and preserved.“

2.2.3 Electronic file backup

Major backup measures reported by survey participants:

1. General strategies:

• Inclusion of ETDs in existing backup plan for digital publications and/or library files
• Use of statewide or national entity for backup (e.g., Florida Center for Library Automation, OhioLINK, National Library of Canada, WVNet)
• Electronic storage at UMI^<3> as a component of the institution‘s backup plan

2. Specific measures taken:

• Regular copying from production server to backup server
• Backup to two off-site locations
• RAID („Redundant Arrays of Independent Disks“) storage
• Weekly or nightly backup to tape (with regular replacement of tapes)
• Off-site storage of tapes
• Backup to CD annually or each semester

Storage of CDs in a safe or fireproof space.

2.2.4 Paper copies

2 institutions are using a paper copy for archival purposes.

4 institutions mention the creation of paper copies for non-archival purposes (as part of the review process or as required by departmental libraries, departments, or committees).

10 institutions make no mention of paper copies.

2.2.5 Physical copies of dissertations

7 institutions require submission to UMI (which generates and retains a microfiche copy of the file) but don't keep a local physical copy.

7 institutions, in addition to requiring submission to UMI, keep a local physical copy or have access to a non-UMI physical copy (5 microfiche, 2 paper).

2 institutions do not require submission to UMI and are not retaining a paper or microfilm copy.

One respondent remarked, „I really view output to microfilm as an enormous plus in terms of digital preservation. That's a luxury we do not have for most of our other digital publications."

2.2.6 Processing of master‘s theses vs. Dissertations

Where both theses and dissertations are processed, 7 institutions process each in the same way for preservation purposes, while 5 have weaker preservation measures in place for theses.

2.2.7 Migration plans

7 institutions reported stronger positions or actions.

6 institutions reported general intentions.

3 institutions had no policies or plans to report.

Migration-related plans include:

1. Making a commitment to future migration of a specific file type or types as necessary
2. charging students $10 per document (over and above UMI processing fee) for future costs to maintain access
3. making an inventory of file formats “as a first step“

delegating migration responsibility to a government-sponsored library agency (e.g., National Library of Canada; interest expressed regarding Florida Center for Library Automation).

2.2.8 Workflow and coordination

Three institutions pointed to the importance of workflow and coordination in managing their ETD program:

„The effort of bringing collections online is integrated into the entire library effort. All departments are involved. Certainly the provost's office and the administration have charged the libraries to preserve the [ETDs]."

There has been a lot of „high-flying rhetoric around ETDs but ours is a very simple system. Others have used our model. Students and faculty understand it. We have a good workflow for doing format checks. People were freaking out. Once they figured it out, it has run smooth as silk."

"The provost has been really instrumental. The archival and administrative issues are important. Support from the top has been vital."

Footnotes:
<1>	The survey was conducted to provide background for a review by North Carolina State University (NCSU) Libraries of its own practices in managing and preserving ETDs. This review was in part prompted by the requirement, established in July 2002, that all NCSU theses and dissertations be submitted in electronic form.
<2>	One institution is not included in this count because in general it does not offer or require master's theses.
<3>	URL: http://www.umi.com/

© This publication and its compilation in form and content is copyrighted. Every realization which is not explicitly allowed by copyright law requires a written agreement. Especially, this holds for reprography and processing / storing by electronic systems.