Libraries of all sizes and types are embracing digital collections, although most libraries will continue to offer both print and digital collections for many years to come. New purchases and purchases of journals, magazines, and abstracting and indexing services are heavily weighted toward digital, while digital books (e-books) are only beginning to become a presence in library collections.
Libraries prefer digital collections for many reasons, including, but not limited to, the following: digital journals can be linked from and to indexing and abstracting databases; access can be from the user’s home, office, or dormitory whether or not the physical library is open; the library can get usage statistics that are not available for print collections; and digital collections save space and are relatively easy to maintain. When total processing and space costs are taken into account, electronic collections may also result in some overall reductions in library costs (Montgomery and King 2002).
Such a dramatic switch from print collections to digital collections has an impact on library users and users’ perceptions of the library. Many researchers have attempted to predict or measure that impact through surveys, transaction log analysis, and other research techniques. Librarians would like to be able to use the information and conclusions generated by the many research studies, especially because it is time consuming to conduct good research on their own and because the best measures of impact come after decisions are already made and collections are converted. Unfortunately, the conclusions of various studies sometimes seem contradictory, and it may be difficult to judge which research studies offer valid and reliable findings. The opinion literature outnumbers the research literature, and it may be a challenge to distinguish fact from opinion.
The purpose of this report for the Council on Library and Information Resources is to help librarians identify reliable research studies, to provide a synopsis of the good studies, and to present an analysis of conclusions. A subtitle of the report might well be the same as the CLIR symposium held March 28, 2003, “What Are Users Telling Us?” Or, “What do user studies tell us about how and why library constituents of all types use digital library resources and are likely to use them in the future?” The goal of this report is to provide information that librarians can use to make important decisions about collections, services, and product design. Also relevant to this topic is CLIR’s January 2002 report “Usage and Usability Assessment: Library Practices and Concerns” by Denise Troll Covey.
Although this introduction refers to the resources as digital resources or digital libraries, the less precise, but more commonly used terms electronic resources or electronic libraries will be used throughout as synonyms.
1.2 Report Outline
Hundreds of recent publications focus on how users interact with or how they feel about electronic library resources. It is important, therefore, to state clear parameters of what is included (and what has been excluded) in this report. Only publications or reports of studies that meet the following parameters are included and analyzed:
- Studies must focus on the use of both electronic resources and libraries (electronic resources through the library, in addition to the library, or in comparison with the library). Studies that are mostly about the Internet, but include a substantive section on libraries (the Pew studies, for example) or those that are mostly about libraries in general, but include a substantive section on digital libraries (the LibQUAL+TM studies, for example) are also included. Internet use studies that do not focus on libraries are excluded.
- Studies or surveys that focus solely on librarians, library staff, library Web sites, or publishers are excluded; only those that study library patrons are included.
- Studies that are limited to the behavior of authors rather than readers are excluded.
- Only research studies are included. Opinion pieces or descriptions of how a library converted their print collections to digital collections are excluded.
- A wide variety of research methods are covered (including surveys, transaction log analysis, experimental). Because different kinds of research methods allow different types of conclusions to be drawn, this report describes the research method used in studies and what types of conclusions made by the researchers are valid in accordance with the method.
- Studies are restricted to those conducted since 1995, or a post-Web world. Some studies compare recent findings with past studies (for example, the Tenopir and King studies), so they may address how usage patterns have changed with the advent of electronic resources, but the main focus remains user behavior in an increasingly digital age.
- Poorly conducted research from which valid conclusions cannot be drawn is excluded.
Applying the foregoing parameters resulted in a pool of more than 200 individual research publications. Some publications describe different phases or parts of large, and often ongoing, research projects. A further distinction was made to separate these large or ongoing studies from the more limited studies and to describe each major study as a whole, rather than as separate publication parts. This led to a distinction between “Tier 1” and “Tier 2” studies.
Tier 1 studies are those major studies that have many publications, sometimes by many different authors. The studies involve hundreds or thousands of subjects over multiple workplaces, work roles, or subject disciplines. Many important conclusions can be drawn from each of these studies and they are typically widely reported and discussed in the library community. Each Tier 1 study is actually a group of studies conducted by a research team. Tier 1 studies are discussed in the greatest detail since they may use multiple methods and provide, at times, complex findings.
The designation as a Tier 1 study was intentionally highly selective. Only eight user studies (actually, groups of studies) were designated as Tier 1 studies, but they represent nearly 100 individual articles or reports. Additionally, nearly that many other publications are designated Tier 2 studies. Tier 2 studies are not of lesser quality than Tier 1 studies; they are just typically smaller in scale or are one-time projects. Tier 2 studies may involve only dozens or hundreds of subjects. They may focus on a single workplace (for example, a single college campus). They provide valuable insights into library user behavior, but are best taken together as a whole to reach general conclusions.
In addition to Tier 1 and Tier 2 studies, selected related materials are briefly described and are included in the bibliography. These include bibliographies of writings about users of digital library materials and several important methods papers.