Understanding the Users of Traditional Archival Collections: Implications for Digital Library Design

The points of departure of this discussion paper are the findings and implications of a study of the users of the National Archives (NARA) that I carried out in 1990 and reported publicly in 1994 (Conway 1994). I interviewed a large (n=800) random cross-section of researcher s, documented the orientation interviews of new and experienced patrons, and picked apart hundreds of reference letters to examine the nature of the written inquiries. The purpose of the study was to inform the design of reference services in a huge, new archival facility at College Park, MD. My two-year project was the first, and only, internal study involving direct interviews with patrons. The insights on the characteristics of the population may be useful in designing digital library systems whose goal, in part, is to help people answer questions that are historical in nature. By historical, I mean questions that depend on chronological time to frame an issue, event, action, circumstance, or process.

Background

In addition to published studies, a handful of significant essays have developed research models that have helped many archival institutions conduct targeted user studies of purely local interest. Elsie Freeman (1984) initially defined the scope of the research needed into patron behavior. I developed a general research model (1986b) and definition of appropriate methodologies. Bill Maher (1986) outlined the administrative value of systematic knowledge of patron use patterns.

Selected Findings (a small sample)

• Patrons approach historical sources with narrowly focused research problems in mind, rather than broad-subject topics. This conclusion appears to hold across types of patron purposes, which range from the purely academic, personal family research, hobbyists, and corporate or governmental clients. This conclusion is reinforced by cross-comparison of queries posed in writing, over the telephone, and in face-to-face interviews.

• Most patrons, by the time they actually show up in an archival facility, are looking for "known items" even if they don't know what those items look like, exactly. Patrons come armed with dates, places, personal names, geographic place, types of documents, and other facets of a "subject" that they are prepared to marshal to improve the precision of their search. Even the most "novice" archival user draws on a storehouse of personal and generalized historical context to frame their queries.

• Patrons are not particularly concerned about the format or media of the information they are seeking. Concrete results are what matter and what drives patron judgments of the retrieval system. Patrons of archives are a perseverant lot, don't expect to be handed answers on a silver platter, tolerate frustrating dead ends and delays, and will be satisfied with a query and retrieval system that gets them close to the evidence they need. Almost everyone "knows it when they've got it."

• When patrons find documents pertaining to their topic, they want hard copies.

• Even as "long ago" as 1990, patrons of NARA were emotionally and experientially equipped to use computers for searching and retrieving information. At least half had specific experience searching in OPACs.

• It is my judgment that very little content learning takes place in archives. What learning that does go on involves mastering a search and retrieval process and "making sense" out of the structures of hierarchically organized document files. Success in archival retrieval (which is analogous to foraging in an enormous, structured, but unknown database) requires that patrons create a mental model of the file structure and then map their specific query onto that model. This process happens over and over again with patrons of varying sophistication, experience, purpose, and expected outcome.

Some Implications

• A "one size fits all" design won't work. The system interface must respond to what it learns about the person using it. Part of what a system needs to learn is the user's skill-level with the system itself, sensitivity to imprecise definition of terms, and facets of the concept of subject, such as time, place, action, participants, outcome, and type of evidence.

• A system should teach users about the mechanics of the"research" process interactively.

• Structure is an important and intuitive navigational mechanism.

• Intellectual discipline (loosely defined) is a valid broad organizing principle. The human impulse to organize what they know and give names to that organization cannot be ignored.

References

Bearman, David. 1989. "User Presentation in Archives." Archives and Museum Informatics 3 (Winter): 3-7.

Conway, Paul. 1986a. "Research in Presidential Libraries: A User Study." Midwestern Archivist 11: 35-56.

Conway, Paul. 1986b. "Facts and Frameworks: An Approach to Studying the Users of Archives." American Archivist 49 (Fall): 393-407.

Conway, Paul. 1994. Partners in Research: Improving Access to the Nation's Archive. Pittsburgh: Archives and Museum Informatics.

Freeman, Elsie. 1984. "In the Eye of the Beholder: Archives Administration from the User's Point of View." American Archivist 47 (Spring): 111-23.

Gordon, Ann. 1992. Using the Nation's Documentary Heritage: The Report of the Historical Documents Study. Washington, D.C.: NHPRC.

Lytle, Richard H. 1980. "Intellectual Access to Archives." American Archivist 43 (Spring & Summer): 64-75, 191-207.

Maher, William. 1986. "The Use of User Studies." Midwestern Archivist 11: 15-26.

Paul, Karen. 1992. The Documentation of Congress. S. Pub. 102-20. Washington, D.C.: USGPO.

Tibbo, Helen. 1993. Abstracting, Information Retrieval and the Humanities: Providing Access to Historical Literature. ACRL Publications in Librarianship, no. 48. Chicago: ALA.