Collaborative Learning Using
Guided Discovery on the INTERNET


Peter Holt, Claude Fontaine, Jane Gismondi, Darlene Ramsden
Centre for Computing Information Systems and Mathematics (CCISM)
Athabasca University
Athabasca, Alberta
Canada, T9S 1A1


Abstract

As computer technology has advanced, information systems have become complex and varied involving more aspects of human capabilities. Despite many recent advances in psychology and human-computer interaction, in information systems design practice has run ahead of ergonomic theory. Professionals need better ways to stay abreast of the most recent empirical and theoretical developments.

We have developed a course in the ergonomics of information systems which is grounded in educational theories of situated learning, collaborative learning, and constructivist (or self-guided) approaches to learning. The delivery platform for the course uses WWW browsers (e.g. NETSCAPE), computer conferencing, and a "talker" facility to support guided access to INTERNET resources and to facilitate collaborative learning in systems development projects. The platform is applicable for education, retraining, professional development, and to support ecologically valid research.


Matching the Domain to the Learning System

Athabasca University is an open university specializing in distance delivery for learners at the time and place of their convenience. In keeping with that philosophy we believe the development of learning systems should be driven by the needs of the learners. Too often these are driven by the technology. Technology is relatively concrete and often captivating; whereas learners' needs and learning theories tend to be abstract and often ill-defined. We have tried to focus the learner's needs in developing a delivery system for a B.Sc. in Computing and Information Systems.

We started with the need for a senior level course in the domain of the Ergonomics of Information Systems. Our approach to developing this course was based on our perception of the rapid development of complexity in this new discipline. As we developed the concept of the course it became evident to us that a good delivery system was based on a model of open learning using guided self-discovery. To illustrate how we reached this conclusion we present a short overview of the evolution of information systems, the development of the discipline of the Ergonomics of Information Systems, and the development of educational and computer-based learning theory.


The Evolution of Information Systems

First let us consider a single universal information system with a series of smaller systems (or sub-systems) including the automated information system, the human being, the organization, and the larger cultural context (see Figure 1). As information systems have developed in the last forty years more aspects of the human being have become involved in the automated information system. The first computers had little impact on information systems. Computers began as back-room curiosities working on obscure mathematical problems for a small group of scientists. Although several visionaries (e.g. Turing) foresaw the multipurpose potential of digital computing, computing remained the domain of mathematically inclined scientists for close to 20 years. At this point there was no ergonomic thrust in information's systems. Human concerns in computing systems were related mainly to speed and reliability of the hardware. With some advances in hardware the convenience of input and output became more of an issue and development efforts focused on increasing the ease of printing and keypunching. Human factors considered were very basic such as the design of keyboards for keypunches and some seating considerations.

[Figures not available in this viewing mode]

After the initial scientific motivation, the expansion of computing was driven primarily by economic factors bounded by technological constraints. Simple business applications became cost effective first. Again, the initial considerations of speed and accuracy fueled the first phase of this expansion to automate organizational information systems. As CRT's replaced batch card entry and teletypes, computers became easier to use. Development of faster CPU's, more software libraries and the beginning of database systems led to rapid automatization of organizational information systems. More members of the organization became involved in the creation, maintenance, and use of automated systems. Ergonomists necessarily became concerned with issues such as continuous hours of work at the CRT and keyboard layouts.

Automatization of information systems functions progressed rapidly in the seventies. Large corporation and government systems became highly automated and this automization spread from mainframes to minicomputers in smaller businesses. Corporations and government agencies became the driving force behind developing the information systems. Database technology developed rapidly and CRT's spread through the desk tops of the organizations and many more types of users were involved in various aspects of information systems. Human factors expanded to include more psychological considerations primarily at the sensory-perceptual level.

The proliferation of PCs, then the MACs, in the 80s added to the complexity of information systems ergonomics. Color screens, iconic interfaces, the mouse, and integrated software paradoxically increased the perceptual/cognitive considerations while making computers easier to use. Management information systems became complemented by Decision Support Systems based on Expert systems technology. Inventory and other commercial systems became common in small business. Libraries became highly auto-mated. As the number of installed units increased the software industry became extremely competitive and users were no longer a captive audience. More attention had to be paid to "user friendliness" and useful "features". Computers became status symbols and advertising appealed directly to the user. Local networks proliferated and the INTERNET expanded further. Users became more vocal in demanding "user friendly" software and the new class of user as consumer became a major force in software development. The affective system became more involved in the interaction with the user. Issues such as gender differences, age differences, and other individual differences became relevant as the range of users of information systems broadened.

This broadening of information systems applications and user groups has continued in the last ten years. Banking machines have become ubiquitous. Applications of computers have spread into every major academic discipline. Computers in education have been espoused by a growing number of mainstream educators. The publishing industry has moved towards more digital multimedia materials. Computer interfaces have gone from primarily text-based with some visual add-ons to include primarily graphical inter-faces with optional sound and digital motion. The competitiveness of the software industry has increased and users expect more from off-the-shelf software and custom coded software in their organization. With the advent of commercial value-added information networks and the INTERNET, there is simply an overwhelming number of information sources and other applications available to almost anybody. While we know of no quantitative measure of information overload, the anecdotal evidence of users feeling overwhelmed by rapid change, new technologies, and new information sources is well established.

Social software in computing is one of the most recent developments with a trend that started with electronic mail and computer conferencing; expanding to include various types of work group software, simultaneous communications systems on INTERNET ( e.g. various chat facilities where users from around the world can exchange ideas and/or entertain themselves); and the development of collaborative systems for education. An example of social software are MOOs (Multiuser object-oriented environments). The simplest of these are simply "talkers". The first talker facilities were designed in 1979 in the UK. This facility allowed a group of people to communicate simultaneously. The first North American talker was introduced in 1991 (Mehlenbacher, B. Hardin, B., Barrett, C. & Clagett, J., 1994) Since that time a variety of these "text-based virtual reality environments" have evolved with the numbers probably in the thousands serving a variety of purposes (e.g. play, education, etc.). Talkers are privately run programs anyone with access to INTERNET can use. A typical talker has a theme (e.g. a space station) which is set by simple text. Usually there are a number of virtual "rooms" that have settings within the general theme. While each talker has a different theme, users basically engage in general conversations. Most of the multiuser communications facilities are recreational but we envision a potential role for education for this technology.

Generally users can talk publicly so all others can hear them or privately to one other user. For example if one were logged on as Bob and said "hello all". The other users would see:

Bob: hello all.

Users can also "emote". In a typical MOO, when user Bob types the command ".emote walks across the room and shakes Joe's hand" other users would see:

"Bob walks across the room and shakes Joe's hand".

The commands available and types and kinds of behavior allowed varies tremendously across these facilities. More complex MOO environments combine virtual objects that users can manipulate along with simple communications facilities. In many ways designing a MOO environment is as complex as designing a real life facility. Since users can "do" things they cannot do in real life and "be" different ages and genders these facilities offer a whole new challenge to social science as a discipline let alone to the field of the ergonomics of information systems.

In summary, societal and organizational automated information systems have become much more complex. Since the advent of computers ergonomic issues have gone from simple sensory and motor human factors to include most aspects of users' affective, cognitive, and social persona (see Figure 2). In this Figure, F1 and Fn refer respectively to sensory, motor, perceptual, cognitive, affective, social, and other factors of a human being's "personality".

[Figures not available in this viewing mode]


Ergonomics

During this rapid evolution of information systems the discipline of the ergonomics of information systems has rapidly changed. At the dawn of the computer age ergonomists were interested primarily in postural issues (seating), motor skills, basic sensory input, and effects of long term viewing of CRTs. As computerized information systems became more complex, more aspects of human capabilities became involved in human computer interactions. Ergonomists turned more to experimental psychology and an information processing view of human capabilities for guidelines to systems design.

Card, Moran, and Newell (1983) in a classic book of readings provided a theoretical framework based on the information processing view of human beings. The basic outline of this theory as applied in human factors is well known and generally involves a linear sequence of information processing from sensory receptors through structures of the brain, to response effectors and resulting action. This approach presented basic design principles based on human limitations in short-term memory, long term memory, and attention. Miller's (1956) found people seem to be able to retain at most seven (plus or minus two) items in mind at a time. Thus inter-faces expecting users to remember more than this number of items are difficult to use. Research on long term memory showed that human memory is not good at retaining lots of specific information all at once. It has to be acquired over time in relatively small increments. Thus complex interfaces should introduce the user to new information in small increments. Attention as a limitation in the human cognitive system has been empirically established (e.g., Neisser, 1967; Lindsay and Norman, 1972; Anderson, 1983). Wickens (1984) expressed the view that the amount that people can attend to at one time may be increased by using different modalities or styles of representation as in multimedia. However, shifting attention takes time and effort and can be disrupting so rapid changes of attention should be kept to a minimum.

Since Card et al's 1982 book several authors have recognized the restrictions of classic information processing theory on human factors or ergonomic approaches to information systems. Chignel (personal communication) argued that the classical processing model represents an incomplete foundation for a human factors account of cognitive tasks as it assumes a human organism that is essentially reactive and does not adequately handle planning, problem solving and other complex tasks (Miller, Galanter, & Pribram, 1961)

Along with development of psychological models of users as a purposive agent, the concept of user as the focus of the design process grew out of practical work in systems development and the field of human computer interaction. Norman and Draper (1984) formalized this approach in the classic "User Centered Design". The book not only focused on user centered design, but broadened the scope of psychological research to more empirical, real world based research. Research on "mental models" (Gentner and Norman, 1979) began to influence psychological theorizing about human computer interactions. This approach focuses not on the "objective" external device or system itself but on the user's model of that device. Mental models and cognitive structures (Payne, 1991) are concerned with the structure of a person's knowledge. In their simplest form, they are beliefs about tasks and systems which guide decisions and behavior. Thus knowledge about mental models can be used to predict behavior. Research discussed in Woods (1984), Norman and Draper (1984), Davies, Lambert, and Findlay (1989) and Allen (1983) suggest that the structure of the presentation on the display screen greatly influences the subject's ideas about the structure and functions of the system. The general finding has been that users' understanding of the system is diminished when the surface representation is extremely cluttered and disorganized and when it does not directly match the functions it is trying to represent. Mental models should conform to the to the actual workings of the system for the users' purposes or show "cognitive compatibility" (Norman, 1988). User interfaces that promote high cognitive compatibility tend to be more usable. Task models should be presented in such a way that their manipulation is con-strained and biased by the structure of relations that actually exist in the represented task or system.

Once designers were interested in higher level cognitive functioning it became important to distinguish between users based on their level of expertise. The usability of a user interface depends to a great extent on the knowledge and experience of the user. Users may be classified into different levels depending on their level of expertise with a software application or user interface. For instance, Schneider (1982) classified users along the following scale: parrot, novice, intermediate, expert, master. Novices know little about the application and they will frequently be reluctant to ask for assistance, since they lack the vocabulary of concepts and terms necessary to express their concerns (Shneiderman, 1987). In current menu systems, novices often benefit the most from the availability of menus (Schwartz and Norman, 1986). In contrast, occasional users will have mastered some aspects of the system, but through infrequent use or practice will sometimes forget key information (Relles, 1981). Occasional users tend to forget the details and are impatient with the need to remember arbitrary syntax and the like. Power users or experts know how to operate the product or system and will know a variety of shortcuts for getting tasks done. In principle one can have a system adapt to the first-time users expertise level. Casual and novice users can have simpler front-ends that restrict their options. Only as their expertise increases will more complex options become available. However, this strategy is complex to implement and can also entail re-learning on the user's part if the system is not well designed. Recently sheer volume of information available to users on the INTERNET has become an issue and there may be a role for "intelligent" agents (Minsky, 1988) that could automate aspects of information discovery, selection, and retrieval for the human user helping to organize information and reduce memory load.

As systems have evolved the ergonomic issues in designing those systems have become more complex. Along with older theoretically derived principals in regard to memory limitations and attention, more heuristic design principles have emerged from the mental model literature and user expertise literature. However, these approaches are embryonic in other than screen design while more complex tasks often involve intuition and heuristics as opposed to hard science. Furthermore as software develops, a principle specific to one piece of software tends to become obsolete and there has been a lack of sound general principles for more complex interactions.

Furthermore, early ergonomists considered both the impact on users of computer technology and on how user limitations could be used to constrain and guide the development of better systems. Most of the research on potentially negative impacts focused on the physical issues (postural strain, eye strain), whereas psychological theory tended to be used in the development of easier to use and more effective systems without an explicit reference to potentially psychologically harmful impacts. Now, with more non-captive users that need to be motivated, some recent focus on the psychological stresses of the information age and a keener interest in societal issues, future research (constrained by the pragmatics of the market) may return to a more balanced ergonomic approach to information systems.

Finally, current approaches do not sufficiently address the social and aesthetic aspects of emerging systems. In these cases practice has ran ahead of theory. As user and automated subsystems develop, the researchers interested in ergonomics and human factors have splintered on the issues of field studies versus experimental approaches to applied research (Monk et al, 1993). Also many soft-ware developers have began to depend upon usability labs for lab-testing of software before release. The cost-effectiveness of this approach versus beta-testing is currently under debate. All these approaches assume the generating of new data instead of working strictly from existing theory. In most cases recent empirical findings at best are loosely tied to theory. Systems development efforts could be considered as real-life laboratories for developing new principles of human psychology.

The job of the information systems designer interested in the ergonomics of the emerging information systems is to keep abreast of the state of theory but also with the rapidly emerging new empirical results in the field. This means the ergonomist needs basic research skill for assessing research results and empirical findings, designing assessments and evaluations, and recognizing relevant variables. Furthermore, the ergonomist needs to be cognizant of the most recent developments in rapid prototyping. Much of this material is still obtained from live conferences and print-based journals but much material also is available on INTERNET. For example, in the overlapping field of Human-Computer Interaction there are very current bibliographies, FAQ (frequently asked questions), and even textbooks on-line at Ohio State University and the University of Colorado. The ergonomist must be able to use the advanced technology of emerging information systems to keep abreast of the field. Finally systems development is generally a team effort. Ergonomists must learn to collaborate not only in systems design but in researching the relevant issues. Given these requirements we looked at educational and computer based learning theory to determine how best to meet our students' requirements. We found what we believe a practical solution in a combination of open learning and computer mediated communications.


Open Learning Systems and Computer Based Communications

In the last ten years there has been a dramatic shift in the perspective of researchers in the area of applying computer technology to education. The focus was once primarily on leading a novice learner through a series of appropriate exercises that gradually induced that expertise in the learner by way of an adaptive program from a stored body of domain experience. (Clancy, 1993). According to Pea (1992) this view of education is based on a transmission view of communication which should be replaced by a "transformative view" in which both the learner and tutor learn and are transformed by their interaction. Furthermore, there is renewed interest in learning as a collaborative process between peers (Pea, 1993). While much research has investigated the computer as a potential collaborator (Collins and Brown, 1989), there has been a renewed interest in human collaboration with the machine as a mediating agent (Harasim, 1992). In fact Baecker (1991) asserts that this new paradigm extends beyond educational technology to the entire field of computer science.

Along with the shift towards the facilitation of human collaboration as the main function of educational technology, advocates of "situated learning" (Brown, Collins, and Duguid, 1989) have promoted moving educational practice into realistic day-to-day settings. Such a strategy is particularly amenable to computer mediated implementation when the discipline area is based on automated tools.

Simultaneous with this change in the theoretical approach to learning systems a number of applied factors have become important in designing learning systems. In the information age there is a need for ongoing education and retraining while global restructuring has reduced the resources available for supporting traditional educational institutions. Industry and students have also pushed for more applied education that leads directly to employment Fortunately, this need not imply very narrow training necessitating constant re-training of the learner. As much as possible the learner should acquire the skills to re-educate and retrain herself at the time and place of her convenience (e.g. home or workplace). While the workplace may make a real contribution to applied education, a broad education integrated into the workplace has much to offer society.

Open learning (Race, 1994) is a solution to many of society's educational needs. In the past this approach has been of limited effectiveness due to separation in space and time of the learners. Collaborative efforts have been particularly restricted and students can be overwhelmed by the wealth of information in a new area of endeavor. However, computer technology can provide a new cost-effective open learning system. Structured hypertext allows students to learn at their own pace but can give guidance similar to an apprenticeship model of learning (Collins and Brown, 1989). Computer technology can remove constraints of distance to support synchronous group activities in a "virtual classroom". (Pea and Gomez, 1992). Pea (1993) claims a mixture of computer technologies can support transformative communications and various forms of group work in which learners and experts learn even at a distance.

However, simply replicating the classroom in "virtual reality" is not a panacea for education. The classroom itself is an old metaphor from an oral tradition and should not be automatically carried over into a new media once again. Even in traditional universities a well-designed text has obviated the need for many aspects of the old lecture and promoted more problem-solving, seminar-like approaches by innovative lecturers. For many learners with heavy work and home demands even a virtual classroom is inconvenience and an asynchronous system with learning at the place and time of their own convenience is ideal (Race, 1994). Li and Mantei (1992) maintain that a great deal of collaborative work is done in the context of casual unscheduled interactions. Consequently a system of guided self-learning supported by asynchronous communications (Harrasim, 1992) with some synchronous casual support ("virtual hallway" chats to partially compensate for the absence of "real world" casual meetings) may meet their need best. For rap-idly evolving disciplines, a computer-based learning system combined with structured hypertext can provide rapid access to the most recent empirical and theoretical results. Together these advantages suggest open learning with computer mediated communications is a pragmatic approach for a course in the ergonomics of information systems.


COMP482: The Ergonomics of Information Systems

Goal and Premises

The goal of this course is to have the student carry out guided research. The produce of the research should be a needs analysis, a systems analysis, an experimental design, a detailed systems design, an experiment, or a prototype system. The six major premises of this delivery platform are: 1. ergonomists and designers of information systems require rapid access to the latest information and knowledge, 2. the learner should not be taught but provided with guidance and new technologies to learn, 3. Learners need a structure to guide them in their learning process, 4. Learning should be situated within the context of real projects and real tools, 5, learning should be a collaborative process facilitated by computer mediated communications wherever possible, and 6. Learning should be able to occur at the time and place of the learner's choice.

Delivery Platform

Over the years we have noticed that technology specific to learning systems soon is made obsolete by mainstream software and hard-ware developments so we focused on available INTERNET and conferencing tools in designing our system. The delivery system for the Ergonomics of Information Systems consists of a number of existing components: a WWW browser a graphical interface tool for connecting to World Wide Web and displaying HTML hypermedia documents, a computer conferencing system for asynchronous group communications, standard INTERNET facilities such as FTP, INTERNET news, a MOO or chat facility for hallway chats and live discussion of pre-posted materials, electronic mail for one to one communication, utilities and software development tools.

The primary delivery vehicle for the course is a hypermedia INTERNET browser such as NETSCAPE. Course materials were created graphics tools and an HTML Editor. Each concept in the course is covered in a one or two pages at most so that the student can exit from on-line materials after completing a concept without having to scroll through pages. Each page leads to another set of concepts through hypertext links which at that level will also be no more than a single page in length. There are links between concepts at the same level. The bottom level of the hierarchy is be a set of bibliographies for the various concepts. Where these bibliographies are on-line on the INTERNET students clicking on a "hot button" are taken directly to the INTERNET site. Thus the course can be viewed as a guide to the underlying bibliographical materials. The guide provides the necessary information for students to understand and use the bibliographical substrate. Where the course material is on-line the course can be viewed as a structured window onto INTERNET resources for the ergonomics of information systems. The guide allows students to learn the material at their own pace using real tools so they can eventually generate a project dealing directly with the material they have learned. The hypertext design also allows student to enter electronic mail, computer conferencing, or the talker at appropriate points in the course with a simple click on a button. Enhancement of these environments are part of the students potential tasks.

The talker facility provides the students with an on-line synchronous chat facility. This "text-based virtual reality" (Mehlenbacher, et al., 1994) (Mehlenbacher, B. Hardin, B., Barrett, C. & Clagett, J., 1994) has separate rooms for courses and a common room. The course rooms have hours with an instructor present but are always available for students to chat among themselves. Students can speak, indicate simple actions (gestural component), and use a bulletin board. We obtained the source via INTERNET and are adopting room descriptions and commands to suit our educational purpose. Much of the administration is done by student volunteers. The strategy is to start with basic a functionality and add features according to assessment of the learners needs. At the present time this application has no "virtual objects" as there are at "Diversity University", the XEROX PARC experimental MOO. We may add the capability for students to add their own rooms. These are being assessed for their usefulness. This is a low bandwidth character based application but video components could be added as part of the project. In the future we see various sites on the INTERNET hosting conferences and talk facilities in their area of expertise.

Software development tools include public domain and low-cost HTML editors and graphics tools, GNU compilers, and PERL interpreters. So far projects have focused on HTML. We would like to add tools such as peepholes (finding peers and friends on network); groupdraw, groupwrite, and shared databases (e.g. for bibliographies). Their usefulness is being assessed as part of the project.


Learning Process

The course is a topics course with the preliminary focus on human factors computer mediated communication, computer-based learning, hypermedia, and computer mediated group work. Course work is not simply be regurgitating or demonstrating more traditional screen design Students build on skills learned in the CIS program. They use a hypermedia structured introductory guide to locate and obtain their own resource materials, develop their own proposal, and complete their own project. Students are expected to generate an original piece of research or demonstrate an innovative design in relation to human factors. If students complete a system it is guided more in relation to the human factors aspect of its design than in its underlying technical implementation. First the student is introduced to the area of the ergonomics of information systems by an on-line HTML Guidebook guidebook. Based on the interests and background of the current students, the coordinator will assign a specific focus area for which each student develops a project description and a strategy for proceeding with research. In the first year the topic will be design of computer based learning systems. The description and strategy are posted to a computer conference for critique by the instructor and other students. Each student is expected to contribute at least one critique (i.e. critically read and summarize another student's proposal) as part of his or her own course. The student is expected to complete the design of a system, a prototype system, or design and/or complete a relevant ergonomic study. Collaboration and peer tutoring are a required part of the course process.


Applications for an Open Learning System

Although this system was developed originally in response to an instructional design requirement for a senior course in the Ergonomics of Information systems and as a tool for working professionals, we saw applications for re-education and training and for research.

Currently we have three part-time researcher assistants working in this area. One is a home-maker who has been out of the teaching profession for eight years, one is a social science research assistant, and one is an unemployed heavy equipment operator who started with Athabasca University on a Federal training program. They are working on talker and conferencing support for collaborative learning, developing hypermedia materials, and testing WWW delivery. All are co-authors of this paper. A primary intent of this project is to continue to employ research assistants from these backgrounds to demonstrate situated learning in on-the-job training in new technologies. We would like to involve more workers from Federal training and re-training programs.

This system also allows us to do situated research - research based in the real world. Much high technology research in education is based upon laboratory research and toy systems that deal with educational concepts divorced from real students. Unlike those research programs we can do ergonomic and educational research starting with a state of the art working system that will be assessed in order to incorporate emerging technology based on real students' demonstrated constraints and needs. Research areas to be looked at in Athabasca University include group dynamics in collaborative learning, hypertext navigation, assessment of need for "intelligence in the system", and ergonomic concerns (in regard to the effect of the length and intensity of use of the system). Enhancements to the system will be determined by learner's needs. Based on the assessments the researchers can identify how the system can be improved with advanced technologies such as artificial intelligence.


References

Anderson, J.R. (1983). The Architecture of Cognition, Cambridge, MA: Harvard University Press.

Baecker, R. (1991). New paradigms for computing in the nineties. Proceedings of Graphics Interface 1991, 224-231.

Brown, J. & Collins, A. & Duguid, P. (1989). Cognition and the Culture of Learning. Educational Researcher, 18,4, 10-12.

Card, S.K., Moran, T.P., & Newell, A (1983). The Psychology of Human-Computer Interaction. Hillsdale, NJ: Lawrence Erlbaum Associates.

Clancey, W.J. (1993). Guidon manage revisited: A socio-technical systems approach. Journal of Artificial Intelligence in Education, 4(1), 5-34.

Collins, A., & Brown, J. S. (1988). The computer as a tool for learning through reflection learning. In H. Mandl and A. Lesgold (Eds.), Learning Issues for Intelligent Tutoring Systems (pp. 1 - 18), New York, NY: Springer.

Davies, S.P., Lambert, AJ.,& Findlay, J.M. (1989). The effects of the availability of menu information during command learning in a word processing application, Journal of Behavior and Information Technology, 8(2). 135-144.

Gentner, D., & Stevens, A. (Eds.), (1983). Mental models. Hillsdale, NJ: Lawrence Erlbaum Associates.

Harasim, L. (1993). Collaborating in cyberspace: Using computer conferences as a group learning environment. Interactive

Learning Environments, 3,2, 119-130.

Lui, R. & Mantei, M. (1991). New paradigms for computing in the nineties. Proceedings of Graphics Interface 1991, 115-122.

Mehlenbacher, B. & Hardin, B. & Barrett, C. & Clagett, J. (1994). Multi-User domains and Virtual Campuses; Implications for Computer Mediated Collaboration and Technical Communication. SIGDOC 94: The Twelfth Annual International Conference Proceedings. New York, NY: The Association for Computing Machinery (ACM).

Miller, G.A. (1956). The magical number seven plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63, 81-97.

Miller, G.A., Galanter, E., * Pribram, K. (1960). Plans and the Structure of Behavior. New York, N.Y.:

Minsky M. (1988). The Society of Mind, New York, N.Y.: Simon & Schuster.

Monk, A., Naridi, B., Gilbert, N., Mantei, M., McCarthy, J. Mixing oil and water? Ethnography versus experimental psychology in the study of computer-mediated communication (1993). Human Factors in Computing Systems: INTERCHI'93 Conference Proceedings. Reading, MA: Addison Wesley.

Neisser, U. (1967). Cognitive Psychology. New York, NY: Appleton-Century Crofts.

Norman, D.A. (1988). The Psychology of Everyday Things. New York, NY: Basic Books.

Norman, D.A., & Draper S.W. (Eds.) (1984). User centered system Design. Hillsdale, NJ: Lawrence Erlbaum Associates.

Payne, S. J. (1988). Methods and mental models in theories of cognitive skill. In J. Self (Ed.), Artificial intelligence and human learning: Intelligent computer-aided instruction (pp. 69-87). London, UK: Chapman & Hall.

Schwartz J. P. & Norman, A. K. (1986). The Importance of Item Distinctiveness on Performance Using a Menu Selection System Journal of Behavior and Information Technology, 5(2), 173-182.

Soby, M. (1994). CMC limits for education. Distance Education On-line Symposium.

Pea, R. (1993). Seeing what we build together: Distributed multimedia learning environments for transformative communications. The Journal of the Learning Sciences, 3(3), 285-299.

Pea, R, D. & Gomez, L. M. (1992). Distributed multimedia environments: Why and how? Interactive Learning Environments, 2(2), 73-110.

Race, Phil. (1994). The Open Learning Handbook. Hillsdale, NJ: Nichols Publishing Company.

Wells, R. (1992). Computer-mediated Communications for Distance Education: An Interactional Review of Design, Teaching, and Institutional Issues. Pittsburgh, PA: ACSDE.

Woods, D.D., O'Brien, Hanes, L.F., & Salvendy, G. (1987). Human factors challenges in process control: The case of nuclear power plants selected applications of human factors in computer systems. (pp. 1724-1770). In G. Salvendy (Ed.), Handbook of Human Factors. New York, NY: John Wiley & Sons.