Conversational systems for adult education and training

The changing face of adult education and training

New forms of educational and training needs

From the organisational side, there is the growing realisation that the needs of individuals for job related education and development are so varied and personal that it is totally impractical to satisfy them by means of a program of standard course offerings. Furthermore, there is such a proliferation of technical information published, that may (or may not) be of importance to certain employees, that the conventional approaches to the handling of information resources by some form of library and information/abstracting service, are breaking down. It has become necessary for all key employees (technical, executive, etc) to take on more responsibility for keeping themselves up to date on new developments that might affect their field of work.

New forms of educational and training methods

As the trends outlined above expand through the business community, similar trends will be seen in relation to adult education, especially in the growing use of distance education in formal educational institutions. To some extent, similar economic factors may lead to a greater use of distance education and electronic networking as the prime delivery media for certain courses. More ubiquitous, however, will be the use of electronic communication media as support to conventional courses. This will be brought about partly by organisational and pedagogic benefits that such systems can offer conventional courses and partly because it will be seen to be the duty of education to use such systems in order to prepare its graduates for the realities of a workplace where they will be obliged to use them.

Learner directed conversation in distance education

The particular focus of this paper is on the effective implementation of computer based group discussion, or "conversational", methodologies on electronic telecommunications networks. This focus is particularly important, as we know much less about how to converse effectively on electronic networks, than we do about electronic self instruction. There is a long history and fairly developed technology of the design, development and delivery at a distance of self study materials in several different (including electronic) media. There is much less known about the running of effective group discussion sessions at a distance. Such teaching methods as seminars (where a group critiques and comments a prepared paper or presentation) or case studies (where groups exchange ideas on how to explain or deal with a problem situation) are traditionally implemented in small or medium sized groups, led by skilled and experienced "facilitators". Much of the success of these teaching methods is ascribed to the facilitators and the skill with which they focus discussion, guide the approaches adopted by the participants, use the natural group dynamics to stimulate interest, participation and deep involvement, pull together what has been learned in the final debriefing discussion, and so on. Can such participatory discussion methods be effectively orchestrated at a distance? How might this be done?

The answers to such questions are vital if we are to learn just how we can utilise the new electronic communication media for education and training. If we cannot use these media for effective participatory group discussions, then there is a serious limit to the extent to which we can use distance education methods effectively. On the other hand, if the economic and other pressures outlined above force our hand into using distance education on a massive scale, then there is a serious limit on the effectiveness of courses in subject areas that are particularly reliant on small group discussion for effective learning.

Although the above mentioned scenario refers particularly strongly to the context of human resources development in the corporate world, the questions are also of importance to the educator, particularly the adult educator. The work which is reported in this paper has been performed to date with groups of students on graduate courses. It will still be necessary to extend this work to the corporate world. However, the basic questions remain the same and the main thrust will continue to be to seek more effective methods for using the new communication media and in the process reduce what might be termed the "loneliness of the long distance student."

Why hypertext for electronic seminars and cases?

Two paradigms compared

Paradigm:	'Instruction' 'Conversation'
OBJECTIVES: (OUTPUT) (why?)	specific pre-defined products standard general negotiable processes variable
MESSAGES: (INPUT) (what?) (when?) (who?) (whom?)	designed pre-prepared instructor one to many created online participants many to many
INTERACTION: (PROCESS) (focus) (analysis) (feedback) (complexity)	behaviours criterion ref corrective one layer thick ideas contents/ structure constructive interwoven layers
DISTANCE EDUCATION: example	Correspondence courses Teleconferencing Video conferencing Computer mediated communication (CMC)

The instructional paradigm is the one that has driven much (though by no means all) of the research and development of the past 30 years that has been performed under the label of educational (or instructional) technology. The conversational paradigm may be seen as the basis of much of the work done on small group work, group dynamics, so called experiential learning and so on. Very often, those working in these areas see little connection between their approaches and methods and those of the instructional technologists. There are of course some very important areas of overlap, for example, the prolific and excellent work on simulations, role play and case study methodology that has been done by educational technologists, but by and large the enthusiasts for one or other of these paradigms have kept apart, professionally speaking.

There is no reason why this should be so. Firstly, there is no intention implied of considering these two approaches as mutually exclusive, or that one is "right" and the other "wrong." They are complementary approaches and each has its role to play in educational and in training systems. Secondly, many real life teaching methods may be in part instructional and in part conversational. Thirdly, there are useful theoretical models of the conversational process as a teaching methodology that have originated on the "instructional" side of the artificial "divide" that seems to separate researchers in our field. one such model, or rather theory, is Gordon Pask's "Conversation Theory" (Pask, 1976, 1984). The work described here is based to some extent on this theory. one hoped for outcome of this work may, therefore, be a closer integration of research and development on teaching.

In relation to distance teaching specifically, one may notice at the bottom of Table 1 that the more conventional "study module" or typical correspondence course model may serve as a good example of the instructional paradigm. Teleconferencing, both audio and video based, is on the other hand a good example of the conversational paradigm in action, as also is Computer Conferencing, as it is most commonly practiced. Computer Mediated Communication has, however, a number of guises. When a computer network is used directly between one student and a distant tutor in order to reinforce a set learning task in algebra and to troubleshoot the student's difficulties on the prescribed problem set, the interchange of messages, although constituting a conversation, may well follow the instructional paradigm in most aspects. A student accessing a computer based instruction package on algebra is having a sort of conversation with the package, but most people would not consider this a good example of conversational education.

Hypertext: Instructional and/or conversational

This "versatility" of hypertext is a potential advantage which is an addition to others mentioned in the literature, such as: asynchronous communication, ability to attract more and richer comments and contributions from students, a permanent record of study, including student comments and annotations, non linear study, random access and so on. It is one potential advantage that is little mentioned, as most authors stress either the non-linearity and student directedness of the user process, or the influence of structure inherent in the content on the authoring process. The work reported here is seeking to use the physical characteristics of a hypertext environment to shape the teacher/learner interactions in particular ways.

Problems of structure and control in CMC

The problem of structure is that participants who take part over a long time period by occasionally logging onto an ongoing discussion on electronic mail have difficulty in maintaining an overall view of the content and structure of the previous discussion. The messages are received in chronological order and seldom is it clear to which previous messages they relate, until they have been read. They are then stored in one (also chronologically organised) "notebook," unless the participant decides to do something special like storing different messages in different "topic" files or notebooks.

Existing mainframe based systems do not have very convenient or versatile methods of storing the structure of the incoming stream of messages. Our students, in general, did not develop any personal structure for storing the messages. As a result, they lost the "big picture" of what had been discussed. Several of the students were asked, after the experience, to draw a structured outline of the content of the whole seminar. Each student's memory of the content and structure of what was discussed was different on important points.

The problem of control refers to the ease or difficulty with which the seminar organiser can keep the participants on task. What tends to happen in this form of discussion over time is that most participants respond to messages as they read them, thus extending the discussion on some hot issue that someone else has just mentioned. If this issue happens to deviate from the originally intended discussion topic, the chances are that discussion will, from that point onwards, ignore the formally assigned task and wander ever further into other topics. This phenomenon can of course also be observed in face to face group discussions. Also, it is not always an undesirable thing to let participants have their head in a discussion just to see where it leads. However, if for some reason there is a specific issue that is to be addressed, excessive digression should be controlled by the group leader, facilitator, or by some other means. our observation in the electronic seminar was that it became more difficult to bring discussion back on track than would normally be the case in a conventional meeting. Requests and reminder messages from the distant leader were largely ignored. In the case of the seminar theme of this first experience, the premises of the issue (whether certain new technologies will in fact be used in education) were challenged and this led the discussion from technical to economic questions, to equity in relation of the "haves" to the "have nots" and so on.

Software for maintaining structure and control

CoSy and PARTICIPATE (PARTI for short) are examples of such systems that are now commercially available. We reviewed experience with both these systems and also gained some practical experience of our own. It seems that the participants still do not place their messages very accurately in the most appropriate "space" in the conferencing structure prepared for them. It is necessary to have a moderator, who reads all incoming messages and decides on the most useful place to store them. The resulting structure imposed on the discussion is thus the moderator's. Also the moderating task is very time consuming and must be performed in a regular, disciplined manner if it is not to bring the whole conference or seminar to a halt. It was for these reasons that we turned to Hypercard, to experiment with other ways of storing an electronic seminar discussion in a structured manner.

Seminars and case studies in a hypertext environment

The Hypercard based seminar network

EXHIBIT #1 [ Separate files, pages 510-513 in the original ]

The sample cards in Exhibit #1 are from a seminar based course that ran through one semester. Each participant (there were 7) prepared a position paper on some specific topic of their choice within a wider predefined topic (in fact the same one that was used in the first email conference described above.) Each participant stack therefore contained their own position paper, that would be spread over some six to ten cards, care being taken to separate each important point on a separate card. This ensures that although one card may give birth to links to a bunch of other cards written by other participants, all these cards will be related thematically. In this way, it was possible to automatically build one group developed structure of the whole of the discourse.

A structured case study discussion environment

At a recent conference, Barnes (1990) argued very convincingly that the key to the effectiveness of a case study exercise is the quality of the case discussion, and this depends in turn on the quality of the case discussion leader. One unfortunate result of this is that effective learning from case study exercises is not always achieved. Also it is always instructor intensive and, therefore, relatively expensive. This factor unduly limits the use of the case study method. At the same conference, I presented a viable approach to overcoming the frequent unavailability of a competent case discussion leader, as well as increasing the efficiency of the available leaders, by means of a networked case study discussion environment (Romiszowski, 1990). This is a "hybrid" system that combines some of the best aspects of computer based learning for presentation and initial discussion of the case materials, with some of the best aspects of computer conferencing, to provide opportunities for reflective discussion of the deeper implications of the case, between various participants and a discussion leader who may be separated by both distance and time.

The computer based learning element is constructed according to the principles of the "Structural Communication" methodology of "conversational self instruction" (Hodgson, 1974; Egan, 1972; Romiszowski, 1986; Pusch & Slee, 1990). The computer mediated case discussion elements can be implemented on any existing electronic network capable of ASCII file transfer, and may be set up to operate as a one on one discussion between a participant and the discussion leader, or as an open discussion between the leader and a group of participants.

An example of the case discussion in action is shown in Exhibit #2. The case is concerned with the analysis of the causes of a human performance problem (card 2.1) and the recommendation of appropriate solutions. The participant studies the basic "facts" of the case in a structured environment (cards 2.2 and 2.3). On completing the case study section of the exercise, the student progresses to the Structural Communication section (card 2.4). After reviewing, if necessary, the key features and description of the methodology (card 2.5), the student progresses to a "response matrix" (card 2.6) and is now invited to construct a solution to the problems encountered. This solution is constructed by selecting all the items considered relevant from an extensive (20 or more item) "menu" of plausible courses of action. This allows the participant a great amount of freedom as there are thousands, possibly millions, of different combinations of response components. The computer analyses the response, using a small expert system that replicates the technical expertise of the discussion leader, identifying significant patterns in the response that may be worthy of comment. In this way, the participant is directed to a series of "feedback comment" screens where certain aspects of the case are further discussed and alternative viewpoints and solutions are compared (cards 2.7 and 2.8).

EXHIBIT #2 [ Separate files, pages 514-521 in the original ]

It is at this point that the participant may "interject" in the discussion, with an original comment or a specific question.

This student generated comment is transmitted to a distant discussion leader, and to other participants if so desired. Discussion then continues on the electronic network.

Ongoing research with this system has shown great potential. We have concentrated, to date, on investigating the educational effectiveness of such an approach. Some of our findings are summarised here. Further details have been published elsewhere (Romiszowski, Jost & Chang, 1990; Romiszowski & Corso, 1990).

• A typical case discussion exercise on this "asynchronous" mode, may stretch over a period of two or more weeks. Typical participants, however, will spend a total of some three hours, spread over some five or six logon sessions. The total time devoted to the exercise by the typical student is similar to the time that would be spent in a face to face, "synchronous" classroom based case study exercise.

• However, every student is participating actively during every logon session. Students are not excluded from the discussion by those who tend to dominate any conversation. As a result, the total amount of discussion, as measured by the volume of commentary and the number of different student generated comments, tends to be much greater than is achieved in a conventional classroom setting.

• The quality of the student generated discussion is also superior, as basic questions related to understanding the case and its implications are handled automatically in the initial Computer Based Learning phase. Students use the network to extend the discussion to a deeper level and make more carefully planned comments. In particular, it was noted that students for whom English was a second language were significantly more participative and productive than is typical in classroom based discussions.

• The acceptability of this technique was reflected in student reaction questionnaires, in which over 80% of the respondents expressed preference for the computer based case studies over similar exercises studied in conventional classroom discussions. It is important to note that the respondents were adult students on full time university courses, who had experience of and easy access to both modalities. Their preferences were thus reflective of the intrinsic aspects of studying cases in these two modalities, rather than extrinsic aspects related to scheduling, distance or accessibility problems.

• The efficiency of the technique was reflected in the ability of one case discussion leader to handle the conversation of over 100 students, on three different geographically distant campuses, over a whole semester. This entailed dealing with some 300 individual student messages per week in the initial stages, and a progressively smaller number as the students settled into using the networking system productively, to raise only important and relatively "deep" issues related to interpreting or solving the case.

• The acceptability of the technique to discussion leaders is dependent on their willingness and ability to communicate by means of computer conferencing systems as opposed to face to face. However, in cases where these prerequisites are met, we have observed a high level of satisfaction on the part of the discussion leader, particularly with the enhanced quality of the discussion, the avoidance of discussion of basic questions related to understanding the case material, and the enhanced possibilities for involving all the participants actively in the discussion.

Future directions

Current and future research

• The further investigation of the most effective and efficient approaches to the use of such integrated CAL/CMC systems for learning/teaching. one current project is investigating certain aspects of students' learning style in relation to the effectiveness and acceptability of network based learning. Another is developing enhancements and improvements to the learning environments that we have created so far. A third is investigating the outcomes of the Structural Communication methodology in terms of transfer of newly acquired thinking skills to new problems and contexts. A fourth project is investigating the effects of Structural Communication on convergent and divergent thinking. A fifth project is evaluating the benefits to learning of incorporating deep level, as opposed to a surface level interactivity in the learning experience.

• The implementation of CAL/CMC projects over considerable distances and over significant periods of time, to investigate the cost benefit factor of our approach in real educational and HRD settings. One project will link university students across three continents in collaborative conversational learning, through electronic seminars and case studying exercises. Another will supply an online learning service to personnel in a corporate setting. These projects are developing special software to link Macintosh and PC based workstations into collaborative "virtual communities" via existing mainframe based messaging networks such as Bitnet and Internet.

Future directions in training and development

The trends also suggest that as business success increasingly depends on so called "knowledge work", the workforce will increasingly depend upon the abilities of locating existing knowledge, evaluating its relevance to the work in progress, and using relevant knowledge creatively to solve the problems or further the goals of the enterprise. This implies that the importance of learning specific content as part of initial training will diminish, as the "shelf life" of the content diminishes, but the importance of developing creative thinking and problem solving skills will increase. Therefore, a further implication is that teaching techniques that are based on the "conversational" paradigm, such as small group seminars or case studies, should be increasingly important in future adult education and training. Increasingly, these educational conversations may take place between small groups of geographically scattered individuals.

Taking these two trends and all their implications, together, we may expect to see increasing use of computers and networks for the delivery of "conversational" as well as "instructional" courses. The totally electronic vocational training system may only be a year or two away from reality. At this time, we are not so skilled at designing effective "conversational" courses for delivery at a distance. Nor have we given much thought to the design of the student-system interface for efficient delivery of such courses. Current progress spurred on by the "IMM revolution" should consider these aspects of interfaces and networking.

The importance of the research and development program outlined in this paper is to discover NOW how to best implement "conversational" learning exercises on the networked delivery systems that we may well be using for much of our training and development effort in the FUTURE.

References

Egan, K. (1972). Structural Communication. Belmont, CA: Fearon Publishers.

Eurich, N. P. (1990). The Learning Industry: Education for adult workers. Princeton, NJ: The Carnegie Foundation for the Advancement of Teaching.

Hodgson, A. M. (1974). Structural communication in practice. In A. J. Romiszowski (ed.), APLET Yearbook of Educational and Instructional Technology. London, UK: Kogan Page.

Horn, R. E. (1990). Mapping Hypertext: The Analysis, organization and Display of Knowledge for the Next Generation of On-Line Text and Graphics. Lexington, MA: Lexington Institute.

Paine, N. (1988). Open Learning in Transition: An agenda for action. London: Kogan Page.

Pask, G. (1976). Conversational techniques in the study and practice of education. British Journal of Educational Psychology, 46. Reprinted in J. Hartley and I. K. Davies (eds) (1978), Contributions to an Educational Technology, 2. London: Kogan Page.

Pask, G. (1984). Review of conversation theory and a protologic (or protolanguage). Educational Communication and Technology Journal (ECTJ), 32(1), 3-40.

Pusch, W. S. and Slee, E. J. (1990). Structural communication: A forgotten application of cognitive theory to instruction. Instructional Developments, 1(2). Syracuse, NY: Syracuse University School of Education.

Romiszowski, A. J. (1986). Developing Auto-instructional Materials: From Programmed Texts to CAL and Interactive Video. London, UK: Kogan Page.

Romiszowski, A. J. (1988). The Selection and Use of Instruction Media (2nd Ed.). London: Kogan Page.

Romiszowski, A. J. (1990a). The case study methodology: Interactive media and instructional design. In H. E. Klein (ed), Problem Solving with Cases and Simulations. World Association for Case Method Research and Application (WACRA). Waltham, MA: Bentley College Press.

Romiszowski, A. J. (1990b). The hypertext/hypermedia solution - but what exactly is the problem? In D. Jonassen and H. Mandl (eds), Designing Hypermedia for Learning. Berlin, Germany: Springer-Verlag.

Romiszowski, A. J., Jost, K. L., and Chang, E. (1990). A hypertext approach to structuring computer-mediated seminars and case studies. In D. W. Dalton (ed), Proceedings of the 32nd Annual International Conference of the Association for the Development of Computer-Based Instructional Systems, 419- 434.

Romiszowski, A. J. and Corso, M. (1990). Computer-mediated seminars and case studies: Possible future trends for in service training and development by means of interactive distance education. Paper presented at the 15th World Conference on Distance Education, Caracas, Venezuela.

Romiszowski, A. J. and De Haas, J. (October 1989). Computer mediated communication for instruction: Using email as a seminar. Educational Technology, 29(10), 7-14.

Romiszowski, A. J. and Jost, K. (1989). Computer conferencing and the distance learner: Problems of structure and control. Paper presented at the 1989 Conference on Distance Education, University of Wisconsin.

Schon, D. A. (1983). The Reflective Practitioner. New York: Basic Books, Inc.

Please cite as: Romiszowski, A. (1992). Conversational systems for adult education and training. In Promaco Conventions (Ed.), Proceedings of the International Interactive Multimedia Symposium, 495-521. Perth, Western Australia, 27-31 January. Promaco Conventions. http://www.aset.org.au/confs/iims/1992/romiszowski2.html