We stand at a critical moment in education. Learning is critical for human happiness and progress, even survival, but all around the world learning is in trouble. However, we have the technology to develop an entirely new learning system, for everyone from birth to death. Distance learning will allow us to reach all six billion people on earth. This paper, other papers, and a recent book picture this new system. Creating it will be a major human adventure.
Paradigms
One paradigm of learning has dominated education for many centuries. Learning is viewed as the transfer of information from a knowledgeable individual to a learner. Transfer may be by lecture in a class, by print (book or the World Wide Web), and by video. Current learning however seems trapped by an inescapable reliance on institutions, mostly schools and universities. This paradigm developed when there were far fewer people on earth. It ignores the parts of learning that go beyond information, such as problem solving and creativity.
Another learning paradigm, tutorial learning, has also existed for a very long time. Learning is viewed as a conversational interaction between a skilled tutor and one or a very small group of students. Unlike the classroom, learning proceeds at an individualized pace until the student succeeds at learning. It is clear that this is the better paradigm, producing superior learning. Those who could afford this approach to learning usually choose it. Thus, the wealthy employed tutors to educate their children.
Tutorial learning with computers
For most of society, the tutorial paradigm is too expensive for a system of learning. But the development of interactive technology, particularly the computer, suggests a new range of possibilities. I have been exploring for over thirty years the idea that the computer, suitably programmed, can be the tutor.
This is not an approach depending on Artificial Intelligence, such as software that is said to be intelligent, but rather on how, and by whom, the learning programs are developed. Such a system will lead globally to much better learning at a lower cost.
As an example, I mention a program we developed about a dozen years ago at the University of California, Irvine, part of the Scientific Reasoning Series (marketed by IBM). The first thing presented to the student is a question, how do you find your body temperature. The student then typed an answer to this question, in English. (A French version was also available, and any language was possible.) When a student said that a thermometer was used, several questions inquire how long the thermometer needed to be kept in the mouth. The program looked for responses like 'keep it in longer,' or 'it should stay in your mouth for several minutes.'
The interaction was in the student's native language, and consisted primarily of questions to the student, and student free-form responses. Multiple choice was used very little. Students proceed at an individualized rate until success is attained. For the entire program, this pattern is followed, with student replies no more than twenty seconds apart. So the student did not spend large passive intervals reading or watching video. We call such tutorial learning sequences highly interactive. It should be contrasted with classes and online learning as seen today.
Tutorial learning is adaptive
Computer-based tutorial learning can not compete with a very skilled human tutor. Nevertheless, we claim that tutorial learning can be much superior to that using the passive information transfer paradigm, as in classrooms. The major reason is that such a program can adapt to the needs of each student. It can at each point, by careful questioning, identify learning problems and offer effective aid. The designers have this responsibility.
An important notion is the Vigotsky concept of the 'zone of proximal development'. Questioning can determine the status of the student, and so find what she or he is now ready to learn.
The questions, the analysis of student replies, and the decisions about what to present next all come, in our activities, from the designers. These are excellent teachers in the area, who frequently talk with individual students and so understand what problems the students are likely to have. They work in groups of about four teachers, pooling their knowledge of students. No programming of any type is required for the designers.
A script, on paper or online, records the full details of the design, the basis for implementation. Several stages of evaluation with large numbers of students, and improvement, are essential to assure that learning is adaptive.
A likely possibility for the future is that student input will be through voice. Empirical evidence for the effectiveness of voice input for learning is essential. For widely spoken languages, existing speech recognition software is adequate. For global use, the speech engines need to be extended to many different languages.
We find that several students working together at the computer encourages better learning.
An experiment
Only small amounts of tutorial learning material are available. The results are encouraging, but not enough to justify the major effort I suggest. An extensive well-planned experiment is essential to determine if this approach is as promising as it seems. It would involve both the development of extensive learning units and the careful evaluation of this material with very many students of varying backgrounds and ages.
Several course-length learning segments are desirable. An important area should be learning for young children, a weak area in many parts of the world. Several areas, extending through adult education, are desirable. I have suggested four such segments be developed and used in the following evaluation. We might begin with just one.
These experiments should be international in scope, involving several languages, including some for students living in extreme poverty. Material should be developed and tested in many countries.
After these 'courses' are developed, they should be intensively tested with large numbers of students of all backgrounds and interests. Professional evaluators should plan this study. The computer saves much of the data while students are learning, but human evaluators will also be required. It is important to consider cognitive information, how each student and each type of student learns, and affective information. This last is particularly important in distance learning, in encouraging students to continue to learn. Our experience is that highly interactive units are intrinsically motivating. We require empirical verification in extensive testing, particularly in developing countries.
The final stage of the experiment is the decision as to whether to proceed to full implementation, based on all the information gathered. This should not be the only experiment involving a new structure for global learning. Other proposals should also be considered and tested.
This experiment will be the beginning of extensive new research in learning.
Full implementation
If the experiment is successful, we should proceed to developing many such learning units, covering all learning from birth to old age. This will be a major international activity, over many years, with many development groups involved. It will be desirable to have several courses in an area, to allow choice and comparison. Eventually the material will need to be available in all human languages, with cultural adaptations when necessary.
Each set of materials should have a professional evaluation, with large numbers of students of many backgrounds. The information from all such studies should be gathered and correlated. We should also plan continuing development of new learning segments.
Any country or group can be involved in development. No one is ahead; we all start at ground zero, so problems of equity should not occur. In some ways, the developing areas may be at an advantage, because they have fewer existing educational structures to offer resistance to such a radically new approach. Large potential profits may encourage countries and companies to participate.
Global distribution
Development of extensive learning is not enough. The materials need to be used by all the six billion people on earth. The mechanisms for widespread distribution may differ greatly over time, and in rich and poor segments of society. We need a variety of distribution methods.
In developed countries we have schools and universities (but sometimes weak), and widespread (but not universal) Internet access. The new units could be used both in formal and informal learning environments. There will be an extensive market, including parents that are unsatisfied with their children's educational opportunities. There will be large profits from this market, so commercial groups will be interested in development and distribution. We must be careful, in this distribution, not to foster the current inequalities in education.
In developing countries, the situation is very different. Teachers and schools are much rarer. In spite of large international efforts from governments, UNESCO and The World Bank, these facilities are generally available only for the wealthy or middle class. The prospects for developing enough schools and training enough teachers for everyone are bleak. The education gap between rich and poor is even greater than in developed countries.
Communications also are much weaker in these poor areas. Half the people in the world have never used a telephone. Internet is only a distant dream for most of these humans, available again only for the wealthy or for the governments. Electricity may not be available. So we may need solar powered computers.
Initially likely distribution in poor areas will be through CD-ROM. Soon two-way Internet access through satellite should be widely available. A study done for The World Bank uses a two-stage satellite system model, global and local, to show that as the numbers of students increase this will be the cheapest way to distribute learning. Some of the profits from sales of the new learning units in developed countries might support this effort.
We can also develop special computers, learning appliances, for use in the developing countries. They would support only the tutorial learning material, so they would need much less capabilities than current common personal computers, with a much simpler, almost invisible, operating system. Probably these will cost less than one hundred dollars. Groups of learning appliances could be set up in village, regional, or city centers.
Conclusions
The prospects are very much worth further study. Several international management groups are needed, for the experiment, for full development, and for distribution. Further details are in the references.
We have an exciting time ahead of us!
References
- Bork, Alfred, Bertrand Ibrahim, Alastair Milne, Rika Yoshii, The Irvine-Geneva Course Development System, in R. Aiken (Ed.), Education and Society, Information Processing 2, Elsevier, Holland, 1992.
- Bork, Alfred, Interview, EDUCOM Review, July/August 1999.
- Bork, Alfred, Tutorial Learning for the New Century, Journal of Science Education and Technology, March 2001, Volume 10, Number 1.
- Bork, Alfred, What is Needed Effective Learning on the Internet, Educational Technology and Society, 4(3) 2001.
- Bork, Alfred, Four Fictional Views of the Future of Learning, The Internet and Higher Education, 3 (2000).
- Bork, Alfred, Sigrun Gunnarsdottir, Tutorial Learning - Rebuilding our Educational System, Kluwer Academic Publishers, New York, 2001.
Papers are available at http://www.ics.uci.edu/~bork.
About Author:
Alfred Bork is Professor Emeritus of Information and Computer Science
at the University of California, Irvine - http://www.ics.uci.edu/~bork/. He has also been at the Dublin Institute for Advanced Studies, the University of Alaska, Reed College, and Harvard University. He directs the Educational Technology Center, a research and development group in technology-based learning. He is Vice President of A Bork Endeavors.
e-mail: bork@uci.edu.
