Judith J. Senkevitch, Ph.D. and Dietmar Wolfram, Ph.D.
School of Library and Information Science
University of Wisconsin-Milwaukee
During the past decade, the explosive growth of electronic media and resources has revolutionized access to information. The Internet, a vast worldwide matrix of computer networks, has made available a vast assortment of information, including databases, text and multimedia files, discussion groups, and official government documents. Once on the Internet, one can find, for example, current postings of available jobs, upcoming cultural and sports events in Milwaukee, resources for disabled individuals, and pending Federal and State legislation. However, it takes both specific knowledge and appropriate equipment to take advantage of these information resources.
Until recently, this access was limited primarily to academic, government, and industrial users. Thousands of information resources were available to this small percentage of network users largely free of charge. Public access was limited to those having the skills and knowledge to navigate electronic networks and the means to afford the technology. This is changing, as the range of information resources and services expands and becomes more affordable, and as public libraries assume a greater responsibility in making computer-based information services available to the public.
As our society becomes more technologically complex, we depend on timely access to relevant information to compete in a changing global marketplace. Information is an increasingly valuable commodity. Yet, many residents of urban areas, especially the poor, still have only limited access to electronic information resources. Unless all residents of Milwaukee and other urban areas can gain access to electronic networks, we risk becoming a society of information "haves" and "have nots"--with attendant disparities in education, income potential and opportunities.
Just as the range of information resources and services available through electronic networks is expanding, so, too, is the number of Americans seeking access to such information. A national poll conducted in 1990 by the Louis Harris organization showed that two-thirds of those adults polled, including low-income people and minorities, were interested in using "computers to obtain online information from the public library or a nonprofit service." In summarizing the implications of the study's findings for public libraries, Westin and Finger (1991, p. 4-6) urge that we use "electronic access to maximize library resources, to narrow the gap between the information haves and have nots, and to empower minorities . . . and senior citizens." .
Public libraries, with their mandate to provide information services to all, offer logical access sites for citizens who lack the technology and expertise to use electronic information resources (McClure, C. R., Bertot, J. C., & Zweizig, D., 1994). While schools are increasingly able to provide area students the technologies, telecommunications, and skills necessary to use Internet resources, many adult residents of Milwaukee lack the access and skills to use information resources that can help them make informed decisions. Electronic networks provide instant access to an array of legislative and policy information at the state and national level and to rich information resources on topics from health care to career and business planning. Electronic networks also provide access to employment opportunities and help-wanted announcements. In a Milwaukee Sentinel column, Deneen Weinz (1994, p. 40D), a Milwaukee minority entrepreneur specializing in job placement, emphasized the importance to Milwaukee women and minorities of being able to take advantage of "opportunities that exist in the advanced telecommunications arena." She advised minority career seekers that "the advent of the information superhighway . . . will unlock many . . . possibilities."
Two earlier studies assessing the current climate for public library access found that public librarians are enthusiastic about national networking, but frustrated by workloads that leave limited time to learn about new resources (McClure, Ryan & Moen, 1993; and McClure, Bertot, & Zweizig, 1994). A recent study of Internet use by public library patrons showed that both library staff and patrons need assistance if they are to use information resources effectively (McClure, Babcock et al., 1994).
Milwaukee area libraries are now beginning to confront the issues associated with network access and service provision. Beginning in October 1994, the Milwaukee County Federated Library System installed a state of the art computer system that gives Milwaukee County residents direct access to holdings of all the public libraries in the county. Residents can look up materials at computers in the libraries, and can also make dial-up queries from remote computer sites. In addition, the new computerized system supports access to the Internet, and to OmniFest, Milwaukee's community-based information service. Installation of the computers at each of the area libraries was completed during 1995. A number of libraries, including Milwaukee Public Library, planned to begin offering direct access to the Internet and other electronic networks in 1995 and 1996. With the availability of advanced computer services throughout the county's libraries, Milwaukee and its public libraries are positioned on the forefront of public Internet access in the nation.
Milwaukee's commitment to information technology could invigorate the city. The area's public libraries can serve as key access points for adults onto the "information superhighway." However, to take full advantage of these services, both library staff and patrons need specific knowledge about the Internet and how its resources can benefit urban residents.
Our study, undertaken July 1995 through June 1996, examines methods for teaching Milwaukee area adults to use newly available information resources in area public libraries. It also examines which types of information resources are of most use to Milwaukee area adults.
Through this project, we developed and tested a model for teaching Milwaukee area adults to use new information resources available through electronic networks, particularly the Internet. Working with Milwaukee area public libraries that began providing Internet access in 1995, we focused on extending access to adult patrons who had no previous access to information resources and technology. Key research objectives for the project included the following:
To compare the effectiveness of resource-specific (cursory) and generalized (in-depth) approaches to facilitating adult novice user access to the Internet;
To compare the effectiveness of graphical and text-based interfaces in facilitating adult novice user access to the Internet;
To study user perceptions of the utility of specific information resources to urban adults;
To develop a model to teach Milwaukee area and other urban adults to use new information resources; and
To identify electronic information resources of greatest use to Milwaukee area adults, including low-income and disabled people.
Newer interfaces to Internet resources such as Netscape, a graphically-based "browser" for the World Wide Web, present users with a wide range of visual input and can integrate pictures, or graphics, with text. It is generally believed that the enhanced visuals --including images with text--give users a more interesting and understandable environment in which to search for information. In order to provide a graphical interface to the Internet, a powerful microcomputer with specialized software and high speed telecommunications connection is needed. The technology requirements for simple textual access are more modest. We also wanted to find out whether longer training periods for new Internet users would significantly improve their effectiveness in using resources.
We divided the year-long study into five stages: information gathering; model development; model testing; evaluation; and model application.
Information gathering. Through interviews with local librarians, we assessed the current level of Internet access in Milwaukee area public libraries. Also, in consultation with librarians, we developed a selected list of electronic services and resources for adults in urban environments. The study focused on adult (rather than juvenile) uses of electronic network resources, since adults are less likely to have had previous access or training.
Model development. We compared the effects of two key elements--type of interface and length of training--on introducing adult new users to electronic services. We tested four approaches, or "models." One model emphasized specificity in the range of resources and services, providing very fast working familiarity with those tools. Participants in these groups received four 90-minute training sessions. A second model represented a generalized approach, providing background on which to build a comprehensive understanding of resources and their relationships. Participants in these groups received eight 120-minute training sessions. In addition, they explored both text-based and graphical interface approaches. (Examples of how the computer screen would appear using a graphical interface and a textual interface appear in Figure 1 and Figure 2 respectively.)
Figure 2. Sample Text-Based Interface.
Each model incorporated a set of guided readings and instructional components, in which a research assistant explained Internet tools and demonstrated their use.
A range of information finding tools and resources were explained in the sessions. These included:
Telnet (or remote login): a tool that allows Internet users to connect to other computer systems;
Hytelnet: menu-based software that lists thousands of categorized resources to which users may telnet;
File Transfer Protocol (FTP): a tool that allows users to copy files from other computers;
Archie: a search tool to find files worldwide that may be copied;
Wide Area Information Servers (WAIS): a listing of databases that users may search;
Gopher: a menu-based tool that contains a variety of local information and allows users to access information from other gopher systems worldwide;
Veronica: a search tool to help locate information on gopher systems;
World Wide Web(WWW): a highly popular hypertext-based tool for accessing a range of information around the world;
Yahoo: a subject-oriented access system that lists information available on the World Wide Web; and
Lycos: another popular searching tool for finding information on the World Wide Web.
Model testing. Based upon Internet availability and interest expressed by librarians, we selected as test sites the Milwaukee Public Library (Central Library, along with representatives from its twelve branch libraries), North Shore Public Library, and Waukesha Public Library. Librarians or skilled library staff participated in this phase of the study; in each of the four test groups they numbered from eleven to twelve people. At the outset of the study, we assessed their level of computer and Internet knowledge and assigned them to test groups to achieve a representative balance of skills within each group. We excluded from the study individuals already very familiar with Internet resources. During the study, research staff monitored each participant for progress, comprehension, and perception of how different networked resources are used. Participants received materials and help-sheets for different tools.
Evaluation. After completion of the test sessions, we conducted a two-part evaluation to assess the effectiveness of each approach. First, at the end of the test sessions, we gave participants a series of questions to answer, using Internet resources, and noted how much time they devoted to each question. We also asked that they complete a questionnaire about their perceptions of Internet resources and their usefulness. The response rate to this initial evaluation was 87 percent (39 of 45 participants). (Reasons for non-response included sick leave and transfers.) By assigning the same specific tasks to each group, we could determine whether any significant differences existed among the groups in the success rate or speed for finding answers. Based on these findings, we created a revised model that provides the most effective method for facilitating librarian and patron access to networked resources.
Second, we conducted a follow-up assessment four months after the end of the test sessions to determine participants' perceived confidence with Internet resources presented in the initial sessions, and their subsequent use of the Internet. This produced a 93 percent response rate (42 of 45 participants).
Model application. Following the librarian test sessions, we extended the research to include adult patrons at North Shore Public Library. (Due to limited public access to the Internet for most of the public libraries participating in the study, potential security concerns, and the implications of the indecency liabilities of the Federal Telecommunications Act then in effect, other libraries were not ready to participate in this phase of the study within the project time frame.) During May and June 1996 we developed and carried out public sessions demonstrating the Internet and its applications, and sessions with a particular focus (such as "job seeking"). Examples of session topics include an introduction to the Internet, using OmniFest, and using the World Wide Web via a graphical interface provided by Netscape software. Sessions were conducted initially by project and library staff, then implemented successfully by the North Shore Public Library staff. While limitations of the project time period did not allow extensive evaluation of this phase, initial assessments suggested growing demand by the Library's public. Initial evaluations also suggested that the opportunity for Internet access may attract new adult users to the library.
Analysis and Findings
One research question of interest to trainers in the field of electronic information services is this: Does the type of interface, graphical or textual, affect user performance? The common belief is that a graphical environment, with its wider range in the types of information displayed (text and pictures primarily) gives users a richer environment that is easier to use. After analyzing the data from four groups of users, using the training approaches described above, we found no significant difference in how well users performed for each interface. We based our conclusions on how well users performed on a series of five questions for which answers could be found on the Internet.
A second question of primary interest to the researchers was whether the duration of the training (short vs. longer time) would significantly affect user performance. Clearly the more training users receive, the better they can use the technology. However, how much is enough? In the initial evaluation of user performance, we found that users in the longer-term training did significantly better.
Are there any significant differences in the time taken to answer each question correctly based on the type of interface or the duration of the training? Here, the length of training did not make a difference. That is, those people who found correct information did so in roughly the same amount of time, no matter how much training they had received. However, the type of interface used did affect the length of time taken to answer questions for two of the assigned questions. Users of the graphical interface took significantly longer for these two questions. This seems puzzling, since we assumed that the graphical interface provided a more user-friendly environment that facilitates locating information more quickly. With a graphical interface, however, users apparently browsed longer before finding answers to the questions, distracted perhaps by the images or by the abundance of information sites. Also, since graphical interfaces--particularly those with many graphics--contain more data, it may take longer for the images and information to appear on the user's screen.
In addition to how well users performed, we wished to determine if they perceived information resources in different ways, according to the type of training they received. Two questions emerged here. How confident did participants feel about using specific resources? How useful did they perceive these resources to be? The data suggest that for five of the ten information access tools (telnet, hytelnet, gopher, veronica, archie), users trained with the text-based interface were significantly more confident than were users of the graphically-based interface (see Figure 3). We had expected that the graphical interface enhancements would make the tools easier to use, thus inspiring confidence; however, this was not the case.
We also expected confidence to be higher for those who received the longer training, but their confidence was significantly higher with only four of the tools used (see Figure 4). Participant perceptions of the usefulness of each tool were similar, except that those in the cursory group did not perceive the World Wide Web (WWW) and its search services to be as useful as did those in the in-depth group. This suggests that longer training may be necessary for the WWW in order for users to better appreciate its capabilities.
Four months after the end of the test sessions, we conducted a follow-up evaluation to analyze longer-term benefits of the training, and to learn whether differences in attitudes and perceptions changed after users had time to explore the tools on their own. Several research questions were important here. Did the group assignment significantly affect how often the participants used the Internet following the training sessions? Access to a graphical interface during the training sessions did not affect Internet use, nor was training duration a significant factor in determining future use. Although participants in the short-term training groups showed significantly less confidence in using certain tools, they did not seem to be affected by it in their subsequent use of the Internet.
Several months after the training sessions, were there any significant differences in the confidence that participants in the various groups had about using specific resources? Overall confidence continued to be higher with the text-based groups than with the graphical groups, but the differences between the groups were smaller after several months.
Were there any differences among groups in how frequently they used specific resources? Did participants undertake any additional training activities after the sessions? Few differences among the groups were found. The duration findings again support the need for more in-depth training for complex tools, such as the WWW. Participants from in-depth groups were more likely to continue learning on their own. However, the interface used in training had no influence on whether participants chose to pursue further training.
Conclusions and Recommendations
One question that has not been answered in research on use of the Internet is whether a graphical interface significantly affects use effectiveness. Anecdotal evidence has suggested that the enhanced interface provided by a graphical environment facilitates searching. However, based on an analysis of responses to the five questions assigned to the participants, we did not find this to be the case. A more comprehensive set of questions is necessary before broader conclusions may be reached. Due to the busy work schedules of the participants, who used work time to complete the evaluations, we could not pursue this question in more detail. The rapid developments in networking diminish the practical importance of the textual vs. graphical argument, since graphical access is much more broadly available today than in 1994 when we developed the initial study. However, from a research perspective, the topic merits further investigation. The present findings suggest that Internet users should not feel disadvantaged if they have a more basic text-based access, rather than graphical access (e.g., the Lynx browser as opposed to Netscape).
The duration of training was a factor in learning to use resources. It is not surprising that participants in the in-depth groups performed better than those in the cursory groups and that they would feel more confident. However, during the follow-up evaluation, the gap between the two groups closed slightly. The initial advantage of the longer training did increase effectiveness over the short term in answering the assigned questions.
We did not expect to find the significant differences in confidence and perceived utility, demonstrated by participants in the text-based groups who rated higher than those in the graphical groups. We had assumed the graphical interface would provide the more user-friendly search environment, leading to increased confidence. This was not the case for many of the resources in the initial evaluation. The graphical interface may have introduced more complexity into the search process, including interactivity and features unfamiliar to some users (e.g., the graphical Windows environment with its pull-down menus, multiple windows, and heavy reliance on a mouse). Although participants began the training sessions with basic computer skills, those in the graphical groups had not used Windows extensively in their work environments. The textual interface, perhaps, presented a simpler tool to use. Another factor that may account for the differences in confidence using resources is how much practice time and computer access participants had outside the training sessions. Those in the graphical groups generally had less access, because they depended on computers available only at work. Those in the text-based groups required simple VT100 terminal access. Several participants in the text-based groups had computers at home with modem access. This permitted them to practice at home by dialing into their OmniFest or other accounts.
Significant differences between the duration groups for confidence and perceived utility led us to conclude that tools with more complex search requirements (i.e., the World Wide Web and search services such as Yahoo and Lycos) require a longer training period to inspire confidence. This conclusion is echoed in findings from the follow-up questions involving confidence and subsequent use of resources. Again confidence and frequency of use of the more complex WWW ranked significantly lower for the cursory groups.
Based on the study findings, we make the following recommendations for a model for systematically introducing new adult users to the Internet:
Provide longer-term training: it inspires more confidence.
Provide more training for complex tools like the World Wide Web.
Provide adequate time for practice in the job environment.
Provide adequate access for practice in the job environment.
Provide graphical access, if available; if not, text-based access works just as well.
In addition to hands-on training, provide background information on the various tools so that users have a better understanding of the different tools and their relationships to one another.
Products of the study include: a tested model for facilitating use of specialized electronic networked resources of interest to urban adults; and recommendations for information resources of greatest use to urban adults, including economically disadvantaged individuals. Project results are already being used by Milwaukee area public libraries, including North Shore Public Library, to document their requests for support to further enhance telecommunications access to the Internet by library patrons. We are sharing with Milwaukee area public libraries the resource lists of interest to Milwaukee area adults, and the components of the model introducing the Internet to patrons.
Research will continue in two areas: 1) to refine further the model for facilitating Internet use for urban adults new to electronic resources, and 2) to gain additional understanding about how best to introduce Internet access to public library patrons, and about the effects of the changing dynamic of information use on the organization.
Internet access in the Milwaukee area continues to grow at a rapid pace. Two years ago, experts hotly debated the question of graphical vs. textual access and whether one could manage with textual access. With the number of Internet providers in the Milwaukee area now providing relatively inexpensive graphical access to the Internet, this question becomes less important in practical terms, but is still of research interest.
The rapidity of change is evident in the shifting popularity of different Internet tools and services. The World Wide Web is currently the most popular way to access Internet resources using graphical software such as Netscape and Microsoft's Internet Explorer. Many other search tools, both subject-based and keyword-based, have been developed to provide additional information location services. Tools such as gopher and Wide Area Information Servers are now considered antiquated, and are quickly being abandoned in favor of the World Wide Web. Successful training models for the Internet will need to factor in this rapidly changing landscape, and will need to provide users with longer term knowledge that goes beyond the tools of the hour. We will also be investigating this area as part of our future research.
Introducing complex and rapidly changing technology into organizations can impose additional strains on an already pressured staff working with limited time and resources. These issues as well require further study.
Because public libraries are the sole public agencies mandated to provide information services to all citizens, they have a critically important role in ensuring access to electronic information sources for the community. Electronic information resources will not replace other more traditional aspects of library services--such as providing good books to read. Rather, they will enhance greatly our ability to tap into specialized knowledge. By promoting this access to information in public libraries, our community can make efficient use of public funds to ensure that we do not become a citizenry of information "haves" and "have-nots."
Acknowledgment: We most gratefully acknowledge a tremendous debt to the good efforts of all the project participants and to the participating libraries (Milwaukee Public Library--Kate Huston, Director; North Shore Public Library--Richard Nelson, Director; and Waukesha Public Library--Jane Ameel, Director)--without whose support the study would not have been possible.
Ayala, R., & Ayala, M. S. (1994). Report card on public library services to the Latino community. Calexico, CA: San Diego State University, Imperial Valley Campus, REFORMA.
Bertot, J. C., McClure, C. R., & Zweizig, D. L. (1996). The 1996 National survey of public libraries and the Internet: Progress and Issues. Final report. Washington, D.C.: U.S. National Commission on Libraries and Information Science, 1996.
Cisler, S. (1996). Weatherproofing a great, good place. American Libraries (October), 42-46.
Cisler, S., ed. (1994). Ties that bind: building community networks. Cupertino, CA: Apple Conference Center.
D'Elia, G. & Rodger, E. J. (1994). Pubic library roles and patron use: Why patrons use the library. Public Libraries (May/June), 135-144.
Flannery, P. (1993). Public libraries and the Internet/NREN: new challenges, new opportunities. Government Information Quarterly, 10(3), 405-406.
Lago, K. N. (1993). The Internet and the public library: Practical and political realities. Computers in Libraries, 13, 65-70.
McClure, C. R., Babcock, W., Nelson, K., Polly, J. A., & Kankus, S. R. (1994). The Project GAIN report: Connecting rural public libraries to the Internet. Liverpool, NY: NYSERNet.
McClure, C. R., Bertot, J. C., & Zweizig, D. (1994). Public libraries and the Internet: Study results, policy issues, and recommendations. Final report, June, 1994. Washington, D.C.: U.S. National Commission on Libraries and Information Science.
McClure, C. R., Ryan, J., & Moen, W. E. (1993). The role of public libraries in the use of Internet/NREN information services. Library and Information Science Research 15(Winter), 7-34.
Milwaukee Sentinel (September 22, 1994). Libraries may get more computers.
Polly, J. A. (1992). Everything I needed to know about the net, I learned in the public library. In Citizen rights and access to electronic information, (D. J. Reynolds, Ed.), Chicago: Library and Information Technology Association.
Senkevitch, J. J., & Wolfram, D. (1994). Equalizing access to electronic networked resources: A model for rural libraries in the United States. Library Trends 42 (4), 696-710.
St. Lifer, E. & Rogers, M. (1994). NCLIS study indicates 21% of public libraries on the Internet. Library Journal, 119 (June), 16-17.
U.S. Office of Science and Technology, Federal Coordinating Council for Science, Engineering, and Technology (1994). High performance computing and communications: Toward a national information infrastructure. Washington, D.C.: Executive Office of the President.
Weinz, D. (1994). Information highway bill green light for minorities. Milwaukee Sentinel, Monday Business, September 19-25, 40D.
Westin, A. F. & Finger, A. L. (1991). Using the public library in the computer age. Chicago: American Library Association, 4-6.
Wisconsin statewide information technology plan: Forward through information technology (1994). Madison, WI: [Office of the Governor].