The purpose of this project was to study the preservation efforts and concerns of college and research libraries across the nation. The participating institutions were drawn from members of ARL (123 institutions) and the ULG (22 midsize universities), as well as 74 leading liberal arts colleges (all members of the OG) and 20 major non-ARL land grant institutions. Together, these institutions represent a large proportion of the academic libraries-large and small, private and public-that are concerned with preserving important research materials.

The study was conducted in two phases and relied on both quantitative and qualitative data gathering. Phase I consisted of the collection of statistical information and other quantitative data relevant to preservation activity in the 116 libraries representing ULG, OG, and major non-ARL land grant institutions. The survey of these libraries was designed to provide documentation on current preservation efforts that was comparable to the information available for ARL members. The project consultant adapted the ARL Preservation Statistics survey for the target group.5 She augmented the survey instrument to include questions that cover basic statistics on library collections, expenditures, staffing, and service activities to provide an institutional context for evaluating preservation efforts. To minimize the demands on participating institutions, these questions were derived from surveys already conducted by members of the OG and from the ARL Statistics and the ARL Supplementary Statistics surveys, which are national standard surveying instruments used by ARL and The Association of College and Research Libraries.6 Additional questions probed the use of digitization for retrospective conversion of library and archival materials and institutional responses to the growing concerns for digital preservation. These questions were drawn from those used in a recent ARL survey on preservation and digitization (Mohlhenrich 2001). Appendix A contains the questionnaire and the accompanying instructions.

ARL mounted the Phase I (“IMLS”) survey on the Web in January 2002 as part of the ARL Statistics and Measurement Program. By early April, ARL had received 68 completed online forms, a response rate of 59 percent. A review of the sample returns indicated that they were representative of the targeted groups and revealed no evidence of response bias. The greatest response came from the ULG; 17 out of 22 (77 percent) of its member libraries completed the survey. Next came libraries of leading liberal arts colleges; 41 out of 74 of these libraries (55 percent) responded. At the lower end were LG institutions; 10 of 20 institutions solicited (50 percent) responded. This latter group does not have an organization that brings together all LG institutions. This may account for the lower response rates from this category and suggests that group identity may play an important role in securing institutional participation. For instance, the response rate to the 2000-2001 ARL Preservation Statistics Survey was 92 percent. Appendix C lists the institutions responding to the Phase I survey.

The results of this survey, representing for the most part a new kind of information for LG, OG, and ULG institutions, constitute baseline data that is both interesting now and potentially useful in the future for longitudinal comparisons. The degree of interest that these data generate in the field and the amount of interpretation they inspire will indicate whether this data collection process should be repeated at regular intervals in the future.

Phase II of the project focused on obtaining qualitative data to complement the statistical data collected in Phase I. The objective of Phase II was to elicit information on attitudes, opinions, and emotions relating to the topic of preservation in order to gain insight into the quality of programs and the motivations and commanding factors for change. Discussions with library directors, preservation administrators, and other personnel engaged even tangentially in preservation issues and activities helped clarify how people working in libraries think about preservation and carry out preservation-related activities. In brief, the project was designed to address the divide between theorists and practitioners and to listen particularly closely to what practitioners have to say.

The data were gathered by means of daylong site visits from late April to late June 2002 at 20 libraries representative of the four target groups. The libraries were part of institutions that were large, midsize, and small; public and private; distributed across the country; and willing to host a site visit. Table 3 represents the organizational breakdown of institutions that participated in the site visits, and Figure 1 shows their geographic spread. The Advisory Committee chose institutions that represented a variety of needs and circumstances, but also typified the range of groups surveyed; it also remained alert to notable deviations from patterns. In total, 76 interviews were conducted with 111 individuals involved at some level in preservation at their home institutions; of these, 55 came from ARL and ULG libraries and 56 from LG and OG libraries.7


Fig. 1: Geographical Distribution of Institutional Site Visits

Qualitative Methodology

The primary data collected in Phase II consisted of transcripts containing the subjects’ responses to a set of 17 questions (see Appendix B). Interviewers were not aiming for the “truth” of the situation, but for an understanding of the subjects’ assumptions about the institutional settings in which they work. Interviewers gave all individuals equal weight and respected their unique interpretations of the institutional frameworks in which they work. Interviewers concluded that, for the most part, staff at the same institution held a fairly uniform sense of institutional priorities, program status, and needs. The multiplicity of interviewers and interviews contributed to the reliability of findings.

Section VI provides general summaries of the responses to questions posed during the interviews. (A fuller analysis of the data drawn from the 76 submitted reports, each consisting of about nine pages of notes, is available at Dividing the data into two files, one for larger (ARL and ULG) and one for smaller (LG and OG) institutions, the consultant coded all data using the computer program ATLAS to derive a list of topics that emerged from the data themselves.


5 Since 1988/89, ARL has conducted annual surveys of preservation activities. See

6 The ARL Statistics ( have been collected and published annually for ARL members since 1961-1962. The Association of College and Research Libraries, with the permission of ARL, has administered the ARL annual statistics survey to all postsecondary institutions for the past two years. Almost 1,400 institutions reported data for the 1999 survey (

7 The number of interviewees might vary, depending upon whether one counts as a “subject” the few extra people-beyond full-fledged interviewees-who contributed to informal conversations during the interviewers’ visits.