Last update at http://inet.nttam.com : Thu May 4 12:36:19 1995 User-Oriented Listserv Operation: A Case Study of PHNLINK April 28, 1995 Sara Kim Vivian Hon Abstract This paper discusses main principles of user-oriented, user-centered listserv operation from the perspective of user feedback, system enhancement, the roles of administrators, organizational dynamics, and monitoring tools. The paper draws mainly from authors' experiences in managing a listserv project called PHNLINK that has been in operation at the Department of Population, Health, and Nutrition of the World Bank. As the spread of e-mail continues to permeate every facet of our lives, the number of listserv forums is likely to multiply. A study into the usefulness and potentials of listserv forums that can be combined with other Internet tools such as World Wide Web is important as listservs carry unique attributes not replaceable by other tools: the creation of human connectiveness through dialogue and information sharing supported by simple media that are accessible and affordable by the mass of people. Contents: 1 Introduction 2 Characteristics of Listserv 3 Case Study: PHNLINK 4 Conclusion: Principles of User-Oriented Listserv Operation References Author Information 1 Introduction The alphabet and print technology fostered and encouraged a fragmenting process, a process of specialism and of detachment. Electric technology fosters and encourages unification and involvement. It is impossible to understand social and cultural changes without a knowledge of the workings of media (McLuhan & Fiore, 1967). Starting with the explosion in the number of electronic mail (e-mail) users around the globe, the word, "Internet" has become an everyday vocabulary in academia, businesses as well as households. Today, the magic threads of electronic connectivity continue to skillfully knit McLuhan's vision of the "global village" into a fathomable picture. This paper focuses on one of the Internet tools, listserv, an e-mail based communication network widely used by the research communities, K-12 school systems, and the general public for primarily exchanging information and conducting collaborative initiatives among other e-mail users who share similar professional and personal interests. With the rapid proliferation of the World Wide Web (WWW), the question regarding the future of the listserv is whether it will become extinct and buried under the multi-layers of the "advanced" technology featured with multimedia interface and multi-pointer capabilities. Another question is whether listserv will maintain and contribute its unique attribute to other Internet tools such as the WWW--the attribute of creating human connectiveness sustained by a simple medium that is accessible and affordable by the mass of people. If the second question seems viable, it will prove to be a useful exercise to examine the determining factors of successful listserv operations and incorporate these factors into the design, management, and modification not only in the listserv operation cycle but also in the development of other Internet tools that can be enriched by the attributes listserv has to offer. The main argument of this paper is that user-centered content design, user-oriented service delivery to audiences who have diverse technological set-ups, and regular system enhancements based on user feedback in a listserv operation environment are key factors in facilitating the very human connectiveness that the users desire to experience. This paper first examines the overall characteristics of listserv as a communication tool, followed by a case study of PHNLINK listserv system that has been in operation at the World Bank since September, 1993. The case study, a combined effort of the former and current PHNLINK administrators, will cover specific operation components and lessons learned over the 18 months of the PHNLINK evolution process. The authors present a series of principles that foster user-oriented listserv operations drawn from the following sources: our experience in PHNLINK management; other e-mail based studies such as CGNET (Consultative Group on International Agricultural Research) and EIES (Electronic Information Exchange System); lessons from social marketing strategies based on communication and marketing principles; and current research in human-computer interaction, more specifically user interface. 2 Characteristics of Listserv Listserv or computer-based group conferencing is one of the Internet tools that is widely used by e-mail users. Listserv or computer forum is devoted to a particular subject area or topic discussed by participants who are initiated into the group through a subscription process. Listserv is either moderated (each posted message is reviewed before dissemination to the entire subscribers) or unmoderated (every message posted to the forum is automatically dispatched to all subscribers). Moderated forum is often employed to prevent unrelated messages to the main topic of the forum from being sent to every subscriber, thus serves as a useful filter system. Listserv enjoys a number of beneficial features of its base technology, e-mail, including asynchronous communication mode, affordable access, and broadcast capability--all of which serve as crucial building blocks in constructing a social structure in an electronic communication era. Specific features of listserv are discussed below: Subscription and Unsubscription: Listserv enables e-mail users to subscribe or unsubscribe to and from a forum with simple commands. A typical subscription (or unsubscription) process takes place by sending a message to the listserv address of a particular forum and typing: Subscribe (or unsubscribe) Forum title User's full name. Users' requests made directly to the system for subscription or unsubscription are electronically processed without any human intervention by UNIX machines on which listserv software programs operate. Distribution: The central feature of listserv lies in its capability to create and maintain a common distribution list of subscribers, that is accessible by all participants of a forum. Subscription or unsubscription requests instantaneously update the distribution list, therefore, current subscription data are readily available to the administrator of the system. Broadcast is made to the name of the distribution list, which usually coincides with the title of the forum, and all subscribers to the forum automatically receive the information through their individual e-mail system. While the administrator has an instant, direct access to all subscribers, subscribers themselves may or may not enjoy the same access depending on whether the forum is moderated or unmoderated. Archival: Listserv offers a valuable feature of capturing the process of on-going dialogues and system evolution by automatically storing every posted message in an archive which is easily accessible by subscribers at any time. A participant in a forum will first retrieve an index that lists all stored messages (preferably sorted by posted date or topic), identifies a filename of a message or article to be retrieved, and requests the file from the listserv system using the filename. In addition, the administrator can upload relevant materials to the archive and he/she alone has the authority to assign filenames to the archived documents. The archived information helps the designers of the system and moderators of discussion groups to retroactively evaluate the development of a forum and at the same time provides a useful conduit to newcomers to a forum, who can orient themselves to the nature and scope of the discussions-to-date. 3 Case Study: PHNLINK PHNLINK (Population, Health, and Nutrition LINK) was conceived by the Population, Health, and Nutrition (PHN) Department of the World Bank in the summer of 1993 with two objectives: i) to connect the PHN community around the world using a simple tool such as e-mail; and ii) to provide an efficient access to the PHN-related World Bank publications via the Internet. PHNLINK started its operation in September, 1993 with approximately 50 subscribers, mainly in the U.S., who received PHNFLASH weekly newsletter, an e-mail reproduction of departmental newsletters containing informative population, health, and nutrition issues. As of April 1995, the number of PHNFLASH subscribers exceeds 1,400 in 52 countries, with an average of 300 article retrievals per month from the archive. In addition, user feedback is regularly solicited for service enhancement and quality improvement of the newsletter contents. What changed the course of PHNLINK that developed intended and unintended personalities as a system and yielded a multiplier effect in the PHN community around the world in terms of the number of subscribers and level of user participation? We will argue that it was neither the superb technological set up nor the efforts of the administrators alone. Listening to user's feedback echoing through the cyberspace and modifying the system to better meet the users' needs rather than demanding the users to adapt to a rigid system were key ingredients that nurtured the PHNLINK cyber seed to sprout and grow. 3.1 Evolution of PHNLINK: Phase I Main Characteristics: This phase can be characterized as a one-way communication flow, i.e. the administrator of PHNLINK feeds weekly newsletters to the subscribers through the system and processes subscription requests by potential users who were not comfortable subscribing themselves or who experienced technical difficulties. The manual subscriptions by the administrator undermined the automatic subscription feature of the listserv system and tied up the administrator's resources that could have been better utilized in developing other system's features. Little feedback on PHNLINK services was sought from the users during this phase. The archive merely stored back issues of PHNFLASH newsletters, offering little incentives on the part of the users to interact with the PHNLINK system. In short, PHNLINK users remained as passive information recipients. Subscription Base: Subscribers were recruited mainly through librarian network groups, announcements made at conferences related to population, health, and nutrition, and solicitation of e-mail contacts of the World Bank staff. As of January, 1994, there were about 100 subscribers in less than 15 countries. Administrator's Role: Most of the administrator's resources were allocated to gathering newsletter materials for its weekly issuance and providing technical support to users including handling subscriptions on behalf of some of the users - a fairly labor-intensive process originally intended to be substituted by the system. 3.2 Evolution of PHNLINK: Phase II Main Characteristics: Figure 2 captures the changing nature of PHNLINK from the one-way communication flow to a two-way interaction between the users and the PHNLINK system where subscribers are no long passive recipients of centrally disseminated materials but rather active participants in shaping the PHNLINK listserv into a socially accepted meaningful tool. Subscription Base: The subscribers themselves became valuable resources in increasing the subscription base through referrals of their colleagues and students to the PHNLINK services. Communications related to referrals were often forwarded by the subscribers to the administrator. In addition, announcements in journals and other listserv forums (Epidemio-Link, Health Net, etc.) contributed to raising the number of subscribers from 269 in 35 countries as of February, 1994 to approximately 1,400 by April, 1995 in 52 countries. There was also an increasing demand during this phase for network opportunities and information sharing with other PHNLINK subscribers. Distribution of Posted Information: A number of PHNLINK subscribers acted as the main focal point of electronic and non-electronic distribution of PHNFLASH newsletters to those with limited or without e-mail access within their organizations and countries (especially in developing countries). This initiative by the subscribers was also targeted at reducing the costs of receiving the same materials by a number of staff within the same organization. Compilation of the number of indirect PHNFLASH readers is based on individual reports of the subscribers during a given period and therefore, the range of PHNFLASH total readership fluctuates from time to time and from organization to organization. Contribution of Materials by Subscribers: Initially, PHNFLASH weekly newsletter contained excerpts from the existing departmental newsletters. Over time, the format of PHNFLASH drastically altered to include job, seminar, and training announcements submitted by the subscribers, notices to the subscribers on other useful electronic services related to the PHN sector, summaries of new publications, comments by the subscribers on PHNFLASH content, and questionnaires on PHNLINK services. Retrievals from the Archive: PHNFLASH archive began to include publications contributed by major research institutions and NGOs (non-governmental organizations) in addition to storing the back issues of PHNFLASH and departmental publications. What's New in PHNFLASH Archive? became a regular segment in PHNFLASH newsletter, accompanied by instructions for article retrieval every week to help new subscribers fully participate in the retrieval activity. The high retrieval rates in May, 1994 (746 articles) and October, 1995 (516 articles) are due to both increases in the subscription base and a large volume of archived articles. Administrator's Role: The initial role of the administrator as a feeder and a blind caretaker transformed into that of a seeker of users' feedback obtained in the form of questionnaires or direct e-mailing. The main characteristics of the administrator's roles can be described as follows: an active solicitor for publications contributed by the subscribers; a "traffic controller" directing the subscribers to useful electronic and non-electronic resources in the PHN sector; a consolidator or merger of other existing electronic services under the PHNLINK umbrella; an informed decision-maker on resource allocation aided by the newly developed monitoring tools on subscribers' profiles and retrieval activities of the archived articles; and more importantly a vital human link to facilitate a full participation of the users in PHNLINK services through education and to create network opportunities among the subscribers themselves. 3.3 Lessons Learned We have described the development of the PHNLINK system from the perspectives of subscription base, utilization of the archive, administrator's role, and types of interaction between the subscribers and PHNLINK system. Several important lessons as well as challenges emerge from our experience and other e-mail based studies. In her study to isolate variables of human behavior that have some correlation with the ability to search an information retrieval system effectively, Borgman identified three areas that emerged as obstacles to effective on-line searching: mechanical aspects to gain access to the system, conceptual aspects to achieve meaningful results from the system, and some design factors that affect the user's ability to use the system (Borgman, 1987). These three factors also pose important implications for optimizing user's ability to fully participate in PHNLINK services as well as the administrator's ability to serve as a link between the users and the system. First of all, the technical means employed by PHNLINK subscribers are extensive, ranging from fidonet users to users with the direct Internet connections. A large part of the effectiveness of PHNLINK is attributed to the choice of e-mail as the base technology, which can be obtained at a low cost with basic hardware set-up. "Successful operation of computer networks. ...requires that a significant mass of users be reached at the various institutional sites on the network" (Schaefermeyer & Sewell, 1988). However, even at the lowest denominator of the Internet technology, a number of subscribers using fidonet or MCI commercial e-mail network could not enjoy the option of article retrievals from PHNLINK system due to default textual messages embedded in users' messages not recognizable by the PHNLINK system. This problem has not been reconciled as of April, 1995. Secondly, the profile of PHNLINK subscribers in terms of their familiarity with e-mail system is as varied as their technical means. While some subscribers are long-time e-mail users, for others, the PHNLINK system was their very first exposure to the e-mail world and listserv arena. First time rejection by the system due to errors during the subscription or article retrieval process can be a major disincentive for inexperienced e-mail users . "The importance of pre-use motivational and perceptual factors suggests that there may be some important underlying psychological or motivation traits that may predict acceptance of system" (Hiltz, 1984). PHNLINK subscribers receive prompt system's feedback upon sending requests for subscription, unsubscription, and article retrievals. In case of an error, the system generates a message informing the user that an error has occurred, but without providing a solution cue to trigger a retrial by the user. At this point, it is up to the user to pursue the root of the problem or simply give up. In order to reduce the number of the "drop-outs" and capture the potential users who are initially turned away by the system, a monitoring tool was developed to track all contacts made by the e-mail users with the PHNLINK system. This record includes e-mail addresses and one line input typed by the users for the administrator to identify the error. Based on this information, the administrator contacted these users with instructions to the correct the errors and invited them to re-try the subscription or retrieval processes. In an effort to support non-native English subscribers, key documents containing useful instructions for subscription and article retrieval have been translated into French and Spanish and are distributed on-line. While refraining from manually subscribing users experiencing problems or faxing the hard copies of the archived materials from the administrator's end, more efforts have been devoted to work with the users both on-line and off-line to help them get on board. Both administrators believe that familiarity with the PHNLINK system translates into user's ability to participate in other listserv services currently available on hundreds of topics. As the subscription base rapidly expands and article retrieval activities become considerably active, more resources should be committed to providing technical support, which is directly linked with users' participation and interactions vis-¯Žvis the system. Perhaps future designing considerations should include technical features which enable the system to recognize the patterns of frequently committed errors in users' messages and dispatch messages containing corresponding solutions to the users from a set of available troubleshooting options. Finally, the enhancement of PHNLINK largely depends on various issues related to system's design. The biggest drawback of the current system is the inability of the users to submit multi-task requests to the system, i.e. 10 article retrieval requests have to be submitted via 10 separate mail messages instead of one. This is problematic for those who pay for on-line connection time and per character transmission costs. Another area of difficulty experienced by a number of subscribers involves typing the incorrect patterns of upper/lower keys at the time of specifying filenames of articles to be retrieved from the PHNLINK archive. The key sensitivity of the system, while providing the administrator with a variety of filename patterns, has been a stumbling block to experienced and inexperienced users. In addition, the current system does not support foreign language characters, thus leaving useful materials untapped and resulting in manually converting the characters into an ASCII format readable by the system. Also not supported by the system include graphics, maps, and tables. The feasibility of supporting non-textual information in the current system environment should be studied in order to make available the vast amount of useful publications so long as the majority of subscribers choose to use listserv as their primary source of information. Lastly, the "browsability" of the index containing archived materials by employing a simple keyword search capability could improve the current retrieval process. A request for the index currently generates five to six pages of filenames and article titles without categories of topics. A keyword search option would help the subscribers engage in a more targeted information retrieval from the PHNLINK archive, reducing the amount of duplicate information previously obtained. 4. Conclusion: Principles and Challenges of User-Oriented Listserv Operation Several key principles for effective operations of user-oriented listserv projects emerge based on the discussions so far. In addition, literature on social marketing strategy\2 and user interface in the domain of computer-human interaction provide significant insights into the organizational dynamics, user needs assessment, user feedback, and the role of the designer and administrator of a system. The listserv operation principles discussed below are by no means exhaustive, but rather serve as guiding principles in the design and implementation of similar Internet-based projects. Listen to User Feedback and Initiate System Enhancement: Seeking and obtaining user feedback needs to be an integral part of a system evolution as manifested by Lewis: "...the central idea of the 'process-is-paramount' view is that it is more important to worry about how to obtain and act on appropriate feedback on the effectiveness of a design than to worry about what design to start with" (Lewis, 1990). Administrators of the PHNLINK system mainly relied on questionnaires to assess the level of users' satisfaction vis-¯Žvis the quality of technical support, ease of the article retrieval system, and the usefulness of article contents. The survey results constituted important grounds for recommending the development of new technical features as well as for allocating resources to make available materials in great demands by the users. The survey also provided information on users' profiles, areas of their professional interests, and other affiliated network systems that offer similar services as PHNLINK. It also helped administrators understand the distribution of subscribers by sectors (population, health, or nutrition), specific geographical locations (urban or rural), and their subscription status (direct subscribers responsible for local distribution or indirect subscribers). In stressing the importance of gaining insights into the users, Favin and Griffiths stated, "Without it [sensitive insight into current beliefs, practices, and perceptions of the target population], message design is left to chance or to the beliefs and whims of officials. These are frequently irrelevant to those for whom the messages are intended" (Favin & Griffiths, 1992). This point is also resonated by Suri: "...nothing could be less reliable than design decisions based upon a solely intuitive grasp of user needs. For an objective critique of user interface design, the user is an essential partner" (Suri, 1993). Therefore, decisions based on user needs and regular feedback provide valuable inputs for system enhancement and necessary resource allocation on the part of the designers and administrators of network. Develop Monitoring Tools: In addition to understanding who the users are, it is equally important to know how they are using the system in their work. For example, four monitoring tools were developed and incorporated in the management of PHNLINK system: 1) cumulative list of subscribers by geographical and organizational breakdown; 2) weekly report on the aggregation of the number of articles retrieved from PHNLINK archive by filenames and e-mail addresses of subscribers; 3) a weekly error report containing the list of e-mail addresses of subscribers who attempted at retrieving articles; and 4) a weekly summary of new subscribers and unsubscribers. These tools coupled with the surveys discussed above provided essential quantitative and qualitative information for the administrator and management on the overall activities of the network. Monitoring tools may vary depending on the goal of a network and available resources (technical and manpower) but the design rationale behind monitoring tools should be rooted in meeting user needs rather than producing empty statistics. Foster Intra/Inter-Organizational Collaboration: Networking and collaboration with other administrators of similar electronic initiatives and information providers can avoid a replication of identical services, and can lead to an efficient resource sharing. PHNLINK served as an example of intra/inter-organizational collaboration during its design and implementation stages. For example, NGOs such as the Clearinghouse of the American Public Health Association and Opportunities for Micronutrient Intervention are regular contributors of materials to PHNLINK in the areas of mother and child health and micronutrient issues. Another notable example of a joint effort is the Partners for Health, a recently established network by the PHNLINK team and the World Health Organization. Within the World Bank, staff members are occasionally asked to forward their e-mail contacts in the field who could benefit form PHNLINK services and publications from other departments are actively solicited for archival. Administrator as the Vital Link: The critical roles of listserv administrators were raised in the discussion on the second phase of the PHNLINK development, which was characterized as a period of rapidly expanding the subscription base and user participation. In order to serve as a vital human link between the user and the system, listserv administrators must go beyond the realm of technical support to actively seek user feedback, negotiate with system developers to improve technical features that reflect user needs, and keep abreast of changing technology that has implications for improving services. Also, the administrators should follow other relevant electronic and non-electronic resources that may present opportunities for collaboration and resource sharing with other information providers. More difficult challenge for the administrator lies in shaping the listserv as a participatory tool that fosters network opportunities among the users themselves for information sharing as well as collaborative initiatives. PHNLINK surveys revealed users' desire to network with other subscribers but the answers to concerns for unsolicited message traffic remain unresolved. One suggestion may be to create a sub-listserv forum among PHNLINK subscribers who wish to carry out discussions on topics important in their work without involving the entire subscription base. As the spread of e-mail continues to permeate every facet in our lives, the number of listserv forums is likely to multiply. A study into the usefulness and potentials of listserv forums that can be combined with features of other Internet tools, such as the World Wide Web, is important as listservs provide a unique attribute as an information repository and a dialogue facilitator. And this attribute can come alive through a partnership between the users and administrators in creating an exciting and relevant communication tool. References [1] Christine L. Borgman, The Study of User Behavior on Information Retrieval Systems. SIGGUE Outlook, Spring/Summer, 35-48, 1987. [2] A. Dix, J. Finlay, G., Abowd & R. Beale, Human-Computer Interaction. New York: Prentice Hall, 1993. [3] Patricia V. D'Souza, The Use of Electronic Mail as an Instructional Aid: An Exploratory Study. Journal of Computer-Based Instruction, 18(3), 106-110, 1991. [4] Michael Favin & Marcia Griffiths, Social Marketing of Micronutrients in Developing Countries. Opportunities for Micronutrient Interventions (OMNI), 1992. [5] Janet Fulk, Social Construction of Communication Technology. Academy of Management Journal, 36(5), 921-950, 1993. [6] P. Golden, R. Beauclaire, & R. Sussman, Factors Affecting Electronic Mail Use. Computers in Human Behavior, 8, 197-311, 1992. [7] Starr Roxanne Hiltz, Online Communities: A Case Study of the Office of the Future. Norwood: Ablex Publishing Corporation, 1984. [8] Starr Roxanne Hiltz & Murray Turoff, The Network Nation: Human Communication via Computer. Cambridge: The MIT Press, 1993. [9] Taplin Shahnaz, Family Planning Communication Campaigns in Ronald Rice & William Paisley (Eds.), Public Communication Campaigns. London: Sage Publications, 1981. [10] Timothy Leary, The Interpersonal, Interactive, Interdimensional Interface in Brenda Laurel (Ed.), The Art of Human-Computer Interface Design. Reading: Addison-Wesley Publishing Company, Inc., 1990. [11] Clayton H. Lewis, A Research Agenda for the Nineties in Human-Computer Interaction. Human- Computer Interaction, 5, 125-143, 1990. [12] G. Lindsay, K. Novak., S., Ozgediz, & D. Balson, The CGNET Story: A Case Study of International Computer Networking. Ottawa: International Development Research Centre, 1994. [13] Marshall McLuhan & Quentin Fiore, The Medium is the Massage. New York: A Touchstone Book, 1967. [14] M. Schaefermeyer & E. H. Sewell, Jr., Communicating by Electronic Mail. American Behavioral Scientist, 32(2), 112-123, 1988. [15] Abigail Sellen & Anne Nicol. Building User-centered On Line Help in Brenda Laurel (Ed.), The Art of Human-Computer Interface Design. Reading: Addison- Wesley Publishing Company, Inc., 1990. [15] Douglas S. Solomon, A Social Marketing Perspective on Campaigns in Ronald Rice & William Paisley (Eds.), Public Communication Campaigns. London: Sage Publications, 1981. [16] Jane Fulton Suri, User trials for information graphics: replacing designers' assumptions with feedback from users. Information Design Journal 7/2, 99-104, 1993. Author Information Sara Kim is the former PHNLINK administrator at the Department of Population, Health, and Nutrition of the World Bank in Washington, D.C. She is currently a graduate student in Educational Technology at the University of Washington in Seattle, WA, U.S. Vivian Hon is the current PHNLINK Administrator at the Department of Population, Health, and Nutrition of the World Bank in Washington, D.C. She is completing her Ph.D. in economics at the University of North Carolina at Chapel Hill, U.S. 1) The authors wish to acknowledge the vision and leadership of Dr. Jean-Louis Lamboray, the former PHNLINK Advisor, throughout the course of PHNLINK design and implementation. Also, the authors would like to express gratitude to Julie McLaughlin for her invaluable contribution to expanding PHNLINK services, Tom Knoblauch for his technical support and advice, and Peter Heywood, the current PHNLINK Advisor for his generous support. This paper represents the views of the authors and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of the Board of Directors or the countries they represent. 2) Social marketing is essentially a combination of public-health oriented social action programs and commercial distribution devices and marketing systems. (Taplin, 1981). Its core principles drawn from marketing and communication strategies include: 1) designing projects and services that meet the needs of intended beneficiaries; 2) promoting products and services creatively based on insights of the beneficiaries; 3) adopting a long-term approach to project planning; and 4) monitoring and adjusting project based on regular feedback (Favin and Griffiths, 1992). User-Oriented Listserv Operation Proc. INET '95 S. Kim