Web technology in public libraries: findings from research
The Authors
Chandra Prabha, Chandra Prabha (Chandra@oclc.org) is Senior Research Scientist at OCLC Office of Research, Dublin, Ohio, USA.
Raymond Irwin, Raymond Irwin (Irwinr@oclc.org) is Research Assistant, at OCLC Office of Research, Dublin, Ohio, USA.
Acknowledgements
# OCLC Robert Bolander, in his role as research assistant, contributed to identifying the libraries that were sampled. Brian Lavoie, Research Scientist, programmatically harvested public libraries in the sample. Both work in the OCLC Office of Research.
Abstract
Public libraries are an integral part of North American communities. With the wide perception that all information people want is freely available, questions arise about the ways in which public libraries are using Web technology. This article assesses public library accessibility via the Web based on a simple random sample of 189 entries coded as “public library type” in the American Library Directory. An e-mail survey to the heads of public libraries in Winter 2002 augmented the data we collected from an examination of Web sites of the sampled libraries.
Article Type:
Research paper
Keyword(s):
Public libraries; Surveys; Internet.
Journal:
Library Hi Tech
Volume:
21
Number:
1
Year:
2003
pp:
62-69
ISSN:
0737-8831
Millions of all ages and backgrounds are turning to the Web as a first stop for finding information. A national survey of 3,097 users found that the average percentage of respondents who reported using only the Internet (20.3 percent) was twice as large as the average percentage of respondents who reported using only the library (9.7 percent) (Urban Libraries Council, 2000).
Findings like these have had an obvious effect on librarians. In a recent nationwide survey, librarians from academic, school, and public libraries were asked about the significant challenges they face. In overwhelming numbers, librarians noted that the biggest challenge they face is the widespread perception that the Internet can provide everything the patron, the student, or the faculty member wants (Jones E-Global Library, 2001). Librarians who see the Web as a threat to survival recognize that indeed Web-based information providers compete with information services that libraries offer.
On the other hand, Web technology significantly assists public libraries in carrying out their traditional mandate more effectively by helping them cater to the daily information needs of typical public library users. The Web frequently offers quick access to information on virtually any topic, ensuring some level of service to members of almost any interest group.
Its scope enables public librarians to call up instantly documents and facts that previously had been much more difficult and labor-intensive to retrieve in a paper-dominated information space. By its very nature, the Web makes provision of current information far more practical than was possible prior to the emergence of the Web. A visit to the Libweb site (http://sunsite.berkeley.edu/Libweb/Public_main.html), which includes home pages of public libraries from around the country, gives the impression that indeed public libraries have assimilated Web technology.
The differences in the perception of the Web’s impact on libraries in general, and public libraries in particular, raised the question of how public libraries are faring with Web technology. It was unclear to what extent public libraries were making use of the Web to meet customers’ information needs. Historically, emphasis on service, not technology, has been the core strength of public libraries. Use of the Web requires new technical skill sets among staff. Approximately 60 percent of public libraries in the USA serve populations of fewer than 10,000 (Public Library Association, 2001, p. 1). Many of these small public libraries are inadequately staffed and therefore open only a few hours a week.
Moreover, introduction and integration of Web technology-based services invariably require new sources of revenue. With the widespread perception that all the information people want is freely available on the Internet, it is becoming increasingly difficult for libraries to successfully lobby for additional funds specifically devoted to the provision of Web-based library services.
Total public library budgets increased 3.9 percent in 2001. But this increase is lower than the increases of the preceding two years – 6.5 percent in 2000 and 7.5 percent in 1999 (Oder, 2002). Even when public libraries are successful in seeking new and substantial funds, their ability to compete technologically with commercial information providers is unclear. The former relies largely on the largesse of taxpayers, while the latter can tap often-significant pools of venture capital.
Public library structures
Public libraries serve residents of a given community, district, or region and are integral parts of North American communities. In 2001, 97 percent of the population in the USA had access to public library services while 3 percent did not. The American Libraries Directory (2001-2002) reported 9,415 public libraries, 1,358 of which also had branches. The comparable number of public libraries in Canada was 706; of these, 124 also maintained branches.
The legal basis for public libraries is complex and varied. In 1998, 53 percent of the public libraries were part of municipal governments, 12 percent were part of a county or parish, 11 percent were part of nonprofit associations or agency libraries; 8 percent were separate government units known as library districts while 6 percent maintained multi-jurisdictional legal status under intergovernmental agreements. Another 3 percent were part of school districts and yet another 1 percent were part of city governments (National Center for Education Statistics, 2001).
The public library makes its basic collections and services available to the population of its legal service area without charge to individual users. In the USA, 78 percent of public libraries’ operating income comes from local sources, 13 percent comes from state government, 1 percent from the federal government and the remainder from other sources (National Center for Education Statistics, 2001). Internet-related expenses, however, are still under 5 percent of the total budget (Oder, 2002).
State governments have a formal structure to support library cooperation, and under state laws, systems, federations, or cooperatives are legal entities, which enable them to handle public funds in return for providing specific library services (Sager, 1989, pp. 278-81). Three out of four public libraries are members of such structures (National Center for Education Statistics, 2001). Because of the diversity of structure and support of public libraries, whether and to what degree public libraries across North America have assimilated Web technology is a complex issue.
The objective of the research that follows was to assess public libraries’ use of Web technology. The results of research on public library Web site content and services will be reported elsewhere. This paper addresses matters of technical infrastructure. We sought data on:
- Which Web domain spaces do public libraries choose?
- Who is hosting public library sites – the libraries, their parent institutions, or library systems?
- Who typically designs the Web sites – personnel from the library, from parent institutions, or from library systems?
- Who maintains and updates public library Web sites and how often?
- How big are public library sites? Data on these, we reasoned, would help us deduce the extent to which public libraries use Web technology.
Literature review
Much of the writing on public libraries and the Web address “how-to” issues of building and presenting Web-based services. A large segment of this work has concentrated on how public libraries can design effective Web sites. Another recurring theme in the literature is the issue of filtering. Some writers have focused on delivery of services via the Internet, targeting special populations through the Web, digitizing local collections, and the Web as a marketing and public relations tool for public libraries. We did not, however, find any study that systematically assessed the extent of North American public library presences on the Web.
The general issue of public library connectivity to the Internet – that is, terminals within the library that allow patrons Web access – has been explored extensively by scholars and organizations alike. The Public Library Association regularly surveys institutions on Internet access and electronic collections and services, including e-mail reference and non-paper document delivery (Public Library Association, 2001, p. 111). The National Center for Education Statistics (NCES) maintains databases on the “service outlets” and governance structures of public libraries, funding and expenditures for electronic information, number of Internet-connected computers within libraries, and estimates of weekly use of information in electronic formats (National Center for Education Statistics, 2001). In the USA, Bertot and McClure (1998) have examined the impact of the National Information Infrastructure (NII) and other public policy initiatives on public libraries’ connections to the Internet. In a way, the present study is an extension of this work. Our study, however, addresses the accessibility of public libraries via the Web.
The United Kingdom Office of Library and Information Networking (UKOLN) has undertaken the only studies comparable to the one presented here. In December 1995, UKOLN issued the “Library and Information Commission Public Library Internet Survey,” which examined the extent of access, types of connections, Internet service providers (ISPs), and Web services offered by public libraries (Ormes and Dempsey, 1995). Two years later, UKOLN launched Web Watch, a project that examined public library Web sites, among others, and reported on the size of the sites and their domain name usage (Kelly and Peacock, 1999). Our study is the first to conduct research on similar issues for public libraries in the USA and Canada.
Study design
Our research is based on a simple random sample of 200 records coded as “public library type” that were drawn from the American Library Directory (2000). Of these, 11 were in fact not public libraries, leaving us 189. Branches were excluded from the sample. As library branches are not autonomous units, sampling was based on the entry for the main library (administrative unit). Commercial search engines (Alta Vista and Google) were used to identify which of the sampled libraries had a Web presence. “Web accessibility” refers to the presence of at least one public library Web page describing library services. The name of the library and its location were used as search terms. In many cases a large retrieval made it difficult to decide when to stop scanning the list of potential matches; though not exhaustive, considerable effort was made to locate the sampled libraries.
We examined the Web sites of the sampled libraries first in winter 2001 and again in winter 2002. These sites were also harvested mechanically using each library’s URL as a starting point. The harvesting software then followed an iterative process in which the hyper-links were extracted from each harvested Web page, and then identified as pointing to either another Web page on the same site as the starting point, or to a Web page on another site. The automated harvesting processes enabled us to differentiate links external to the library site from links to pages within the site. This was necessary to estimate the size of the sites under study.
Libraries from 44 states, four Canadian provinces, and the District of Columbia were represented in the sample, including 15 from Illinois, 12 from New York, 11 from Texas, and nine from Wisconsin. The sample was regionally diverse, drawing 24 libraries from the West, 19 from the Southwest, 26 from the Southeast, 23 from New England, 51 from the Midwest, 22 from the Great Plains, and 24 from the Mid-Atlantic region. The sample also included libraries of various sizes, measured in terms of population served, annual income, and volumes in collections.
We augmented unobtrusively gathered data with data from an e-mail survey. A total of 120 libraries in the sample had identifiable e-mail addresses; 105 of these were valid. A total of 36 of the survey forms were returned for a 34 percent response rate.
Findings
In winter 2001, 58 percent of the sample was Web accessible; this percentage increased to 77 percent in winter 2002. The findings presented in this article are based on winter 2002 data. We report on which Web domain spaces public libraries are choosing, whether they are hosting the Web sites themselves, the extent of their Web sites, size of entry pages, and whether the libraries are designing and maintaining the sites themselves.
Top level domains public libraries use
Many public libraries tend to express themselves in terms of the populations they serve in a particular geographical location – a city, a county, a state, and the like. Thus the locality where their primary population lives is frequently used in naming public libraries. Public library names such as the Webb City Public Library, Free Library of Springfield Township, and Black Creek Village Library illustrate the point.
Several libraries in the sample that identified themselves by geographical location transferred this identity to their allocated Web space as well. The most popular domains for public library Web sites were related to geographical location (46 percent). The URL http://www.hartshorne.lib.ok.us, for example, reveals that the public library serving the community of Hartshorne is located in Hartshorne, Oklahoma in the USA. Likewise, the library serving the city of Port Moody, British Columbia, Canada is represented by the URL http://www.library. port-moody.bc.ca/.
Our research indicates that many other public libraries see themselves as nonprofit service organizations, without regard to location (Table I). Of the public library home pages, 32 percent had Internet addresses that ended in “.org,” which is typically used for not-for-profit organizations. Another 12 percent of the public libraries indicated use of commercial vendors (.com), while the remaining 8 percent noted the use of a community network (.net), government (.gov), or educational institution (.edu).
Public libraries’ preference for the location domain reflects the practice of naming libraries by the political geography they serve. Likewise, public libraries’ choice of the organization domain reflects the non-profit aspect of public libraries. These two domains together account for 78 percent of all public library Web addresses.
Who hosts public library Web sites?
Web addresses also reflect public libraries’ administrative affiliations. In reality, libraries can be – and often are – members of both a library system/network and a political community. The library in New London, Connecticut, for example, is supported primarily through city taxes and is clearly a substantial information resource for the community. At the same time, in order to provide enhanced service to the citizens of the city, the library’s management chose to join Libraries Online (LION), a non-profit alliance of 22 libraries that share resources and maintain a common OPAC. Its Web presence is associated with the LION site (http://www. lioninc.org), rather than the official site of the city of New London (http://www.ci.new-london.ct.us/) or any other governmental units or ISPs.
In the Web environment, the options for public libraries appear complex. In establishing and maintaining an official presence on the Web, public library directors and boards seem to have had two choices: to connect the institution to a larger entity or stake out their own “ground” on the Web. The majority were associated with other institutions, both within the library realm (regional, state, or national library systems) and outside of it (governments, commercial services, universities, and community networks). A total of 34 percent were connected to non-library entities (see Table II); 28 percent of public libraries were tied to library systems and 38 percent of libraries chose to “be independent”. Put another way, about two-thirds of public libraries on the Web were accessed through the pages of library-based administrative units.
Many non-library entities host public library pages including, most prominently, municipal, county, and state or provincial governments (20 percent). The remainder includes commercial hosting services (9 percent) like Geocities, Angelfire and local ISPs; community networks (3 percent), which are often established by chambers of commerce to promote local trade and tourism; and universities (1 percent).
Size of public library Web sites
One basic question involved the size of public library sites. A reasonable measure of a site’s extent is the number of pages that it contains. Here we used the definitions proposed by O’Neill and Lavoie (2000, pp. 57, 59). They defined a “page” as “a distinct information unit composed of one or more HTTP-accessible files, referenced and accessed in its entirety by a single URL”, and which can consist of a number of files viewed simultaneously in a Web browser when the page’s URL is accessed. Based on computerized internal link counting, we found that public library Web presences ranged from a low of one page to a high of 7,451 pages and that the average numbered slightly less than 305 pages; the median, however, was ten, suggesting a manageable grouping of Web pages for most public libraries.
The size of the entry page varied from 522 to 32,202 bytes, but averaged 4,532 bytes; the median front page was 6,598 bytes. This contrasts with the front pages of popular commercial Web directories/search engines like Yahoo (17,045 bytes), Looksmart (15,771 bytes), Alta Vista (10,156 bytes), Excite (42,107 bytes), HotBot (25,546 bytes), and Lycos (37,578 bytes). (These numbers are based on examinations of the sites in March 2002.) The typical public library front page, then, is about one-fifth the size of the entry page for the average commercial search engine.
Design and maintenance of Web presences
We also wanted to know who does the design work on public library sites. Some library Web pages indicate the designer and the frequency of updates. Most, however, do not. We surveyed directors of libraries in our sample to obtain data about who was responsible for design and structure of Web presences.
Over half (58 percent) responded that library staff did design work in-house, while 11 percent said that the work was done at the system level. Equal proportions (8 percent each) left it to parent governments and commercial services, while 14 percent cited some other arrangement – usually collaborative work among institutions or the use of volunteers from the community.
We also asked public librarians about site maintenance. Almost three-quarters reported undertaking regularly scheduled revisions of their pages. Two-thirds of respondents noted that the work is done within the library, while 19 percent said that updates were done at the library system level. The largest number of respondents (37 percent) reported that revisions were done on a weekly basis; 26 percent updated their pages monthly, 14 percent daily, 11 percent quarterly, and 11 percent twice a year.
Size of the population served matters
Statistics on populations served were available for 159 libraries in the sample; 124 (78 percent) of these libraries had a Web presence and 35 (22 percent) did not. Among the former group, the average population served was 225,809 and the median was 16,270. The average population served for the latter group was 10,202, while the median was 2,100, indicating that public libraries serving smaller populations – and usually operating on more limited tax revenues – were less likely to be “on the Web” than were larger libraries.
Generally speaking, libraries serving larger populations were more likely to host their own Web pages than were libraries serving smaller communities (see Table III). For this analysis, we ordered the libraries by the size of the populations they served, and then divided them into four groups of equal size. Each quartile had 31 libraries. Libraries in the top (first) quartile served populations of more than 42,000; libraries in the second quartile served populations between 42,000 and 16,000; libraries in the third quartile served populations between 15,999 and 6,440; and the fourth (bottom) quartile served populations of fewer than 6,440. Libraries in the top quartile were four times more likely to host their own Web presence than were libraries in the bottom quartile.
Moreover, both the extent of the Web presence (in pages) and the entry point size (in bytes) correlated to public library service populations (Table IV). The smallest group of libraries in our sample had the fewest pages (median 1), while libraries serving the largest populations had the most Web pages (median 101). Likewise, large library front pages contained more information encoded in bytes (median 9,313) than did the smallest libraries (median 5,069).
Conclusions
Our research shows that for most public libraries the technical infrastructure and administrative willingness to build a significant Web presence are in place, especially among larger institutions. Public libraries have taken advantage of all the options available to them. They appear devoted to becoming Web accessible and to providing services to users remotely. In short, public librarians seem to have accepted the challenge of creating a Web presence on the Internet and are using available means to continuously meet the challenge. From our research several conclusions emerge.
First, public library administrators have clearly recognized the importance of having a Web presence. Web presences of the public libraries in our sample increased from 58 percent in 2001 to 77 percent in 2002. Given the anxieties that librarians have expressed elsewhere about the challenge of the Internet, there is no reason to believe that this trend will abate. Libraries are acutely conscious that they must be “on the Web” in order to compete with other suppliers in the delivery of electronic information.
Second, in their movement to the Web, public libraries have not forsaken their basic identities as non-profit information providers to well-defined geographic areas. Nearly half utilized location-indicative domains, while about a third identified themselves as “.orgs,” an extension often used by not-for-profit organizations.
Third, public librarians have been quite resourceful in establishing Web presences. For most, this response comes from necessity arising from a lack of both staff expertise and money. The vast majority of individual libraries in our sample did not host their own Web sites, opting instead for less costly alternatives that require no library staff experience or additional technical training. A total of 62 percent of public libraries “attached” themselves to other entities – systems, parent institutions, community networks and the like – a number borne out by our survey of library administrators, three-quarters of whom reported choosing “outside” hosting options. When prompted for an explanation, nearly half cited cost and more than 60 percent noted lack of technical expertise among the staff. Not surprisingly, for middle-sized and smaller public libraries time, money, and technical skills are key constraints in the building and maintenance of a Web presence.
Most public libraries manage these obstacles with a little assistance from a variety of sources. Among libraries that do not host Web sites on their own machines, a plurality work within the “library world” to reserve a place on the Internet. We found that library systems and state, provincial and national libraries hosted 28 percent of public library Web presences, a number that closely mirrored librarians’ survey responses (31 percent). About one-fifth of the libraries integrated themselves into their parent government Web sites and approximately one in ten chose what those libraries decided was the most economical option – commercial ISPs or community organizations.
Fourth, we found that size of the population the library serves matters. Smaller libraries in our sample were much less likely to have a Web presence than their larger counterparts. Libraries serving larger populations hosted their own Web sites in greater proportions than those serving small communities. Moreover, these larger libraries, which generally enjoy better funding and staffing, put up more Web pages and included more information on their start pages than smaller libraries.
Fifth, though public libraries struggle with funding and staffing issues, they are active in the design, update and maintenance of their sites. Almost seven in ten (69 percent) design their Web presence using existing paid library staff or utilizing library system resources. Moreover, most survey respondents (51 percent) reported updating their pages either daily or weekly, and more than three quarters (77 percent) update at least monthly.
Our study clearly indicates that public libraries are participating in the revolution brought about by the Web. Nonetheless, they face severe competition in the Web world, from commercial search engines and directory services, and even from each other. The Urban Council study referred to earlier lays out three potential scenarios: significant changes in public library missions and services, peaceful coexistence with the Internet for the foreseeable future, or initiation of a period of decline for public libraries. Which scenario will prevail? This is an open question.
Table ITop level domains of public library Web presences
Table IIWho hosts public library Web sites
Table IIIWeb site hosting: libraries hosting on their own vs libraries becoming a node in other hosting entities
Table IVExtent of public library Web presences and size of entry pages by library size
References
(2000), American Library Directory, 52nd ed., R.R. Bowker Company, New York, NY, .
Bertot, J.C., McClure, C. (1998), The 1998 National Survey of US Public Library Outlet Internet Connectivity, www.ala.org/oitp/survey98.pdf, .
Jones E-Global Library (2001), "The role of librarians in the digital age: preliminary findings", www.egloballibrary.com/, .
Kelly, B., Peacock, I. (1999), Web Watching UK Web Communities: Final Report for the Web Watch Project, www.ukoln.ac.uk/web-focus/webwatch/reports/final/rtf-html/report-Appendix.html, .
National Center for Education Statistics (2001), "Library statistics program: public libraries", http://nces.ed.gov/surveys/libraries/public.asp, .
Oder, N. (2002), "The new wariness", Library Journal, Vol. 127 No.1, pp.55-7.
O’Neill, E., Lavoie, B. (2000), "Bibliographic control for the Web", Serials Librarian, Vol. 37 No.2, pp.53-69.
Ormes, S., Dempsey, L. (1995), "Library and Information Commission Public Library Internet Survey", www.ukoln.ac.uk/services/papers/ukoln/ormes-1995-01/, .
Public Library Association (2001), Public Library Data Service Statistical Report 2001, Public Library Association, Chicago, IL, .
Sager, D. (1989), Managing the Public Library, 2nd ed., G.K. Hall and Co., Boston, MA, .
Urban Libraries Council (2000), Impacts of the Internet on Public Library Use, www.urbanlibraries. org/Internet %20Study %20Fact %20Sheet.html; also available at: www.urbanlibraries.org/pdfs/finalulc.pdf, .
Further reading
Moore, K., Knox, K. (2001), "How can we survive in reality library?", Computers in Libraries, Vol. 21 No.10, pp.34-8.
Puacz, J. (2002), "Catching (and keeping!) e-patrons", Computers in Libraries, Vol. 22 No.1, pp.12-14, 71-2.