Agile learning with highly interactive computer modules

Citation

Bork, A. (2002), "Agile learning with highly interactive computer modules", Library Hi Tech News, Vol. 19 No. 4. https://doi.org/10.1108/lhtn.2002.23919daf.001

Agile learning with highly interactive computer modules

Alfred Bork

Education is … the key to … development that is both sustainable and humane, and to peace founded on mutual respect and social judgment. [I]n a world in which creativity and knowledge play an ever greater role, the right to education is nothing less than the right to participate in the life of the modern world (UNESCO, 1996).

… education is an essential human right, a force for social change – and the single most vital element in combating poverty, empowering woman, safeguarding children from exploitative and hazardous labour and sexual exploitation, promoting human rights and democracy, protecting the environment and controlling population growth. Education is a path toward international peace and security – 130 million children in the developing world are denied this right [education] – almost two thirds of them girls. Nearly 1 billion people are illiterate – the majority of them women (Annan, 1999).

While working on this paper I attended a series of lectures at a national meeting of the National Academy of Engineering. Seeing the topics in advance, I had no idea that I would hear anything that was related to improving learning, my major interest.

But I did. One of the speakers was Michael Molnar, with Cummins Engine Company. His topic was Agile Manufacturing, a subject unknown to me. He started with an example of a manufacturer of poker tables. The purchaser had great control over the individual table. He estimated that there were over five million different combinations. Clearly the manufacturer could not stock all these, so they had to be manufactured when ordered. His second example was from his own company, Cummins, a manufacturer of diesel engines. Almost every engine it delivers is unique of needs to the customer. The problem of agile engineering is how to organize production so that such extreme flexibility is possible at reasonable costs.

I could not help but compare this with learning. I thought particularly about the large beginning courses in almost all universities; similar considerations would apply everywhere in learning environments. We certainly do not deliver in education agile products to our clients. That is, we do not tailor the learning experience, in traditional environments, to each student. Rather we have a one size fits all strategy, with minor variants. We are not responsive to individual student needs. This essay explores how to create an agile learning system, and deliver a unique learning product to meet each student's needs, at reasonable costs, for the future of education. The papers and books referenced at the end give far more details, including information on production, delivery, and large-scale development.

Learning problems

The major problems in learning today are noted by five points:

1. 1.

   Quality problems. Many students today at all levels do not learn. Our whole system of grading in schools and universities is based on the notion that not everyone will learn, or that not everyone will learn well. We focus on the few students who fully succeed, and in effect ignore most of the others. From the manufacturing point of view, we claim success by pointing out these successful students and ignoring the failures. Yet it was established many years ago by experiments done by Benjamin Bloom and students in the Chicago public schools system that everyone can learn, given the learning environment appropriate to each individual student.

2. 2.

   Paradigm problems. Our current dominant paradigm for learning views learning as a transfer of knowledge from an individual or a print document to a student. Indeed we focus almost entirely in talking about learning on knowledge and information. This is no surprise. This paradigm ignores many important features of learning, such as problem solving, socialization, and the cognitive process. It may not be the best paradigm for learning.

3. 3.

   Lifelong problems. Learning begins at birth, and continues throughout life. But we tend to view learning as a question of schools, universities, or training. Given the rapid change of the modern world we need to consider learning as a life-long process. Presently we have major articulation problems between different phases of this process, such as the phase between speaking and reading for a young child, or the phase between secondary school and university. We need to view learning as lifelong.

4. 4.

   Further, our learning processes today often do not encourage lifelong learning. Many students find learning boring and unpleasant, and so have no interest in continuing it.

5. 5.

   Global problems. Many aspects of learning will never be effective unless they occur globally, for everyone on earth. For example, we cannot overcome the problems of terror and violence by educating those in only one country in non-violent directions. So we can expect to have terrorists continue to plague us in the future. We cannot expect to have a world free of poverty, to use the World Bank phrase, unless everyone has equal learning opportunities. Education, lifelong, should be considered a right of all people, not just the favored few.

6. 6.

   Economic problems. To some extent this is related to the previous problems, particularly in the very poor parts of the world. Learning today costs too much. But even in wealthy countries the costs of education are often beyond the means of individuals and governments, particularly we consider lifelong education. Education should be affordable by all.

How can we overcome these problems?

I consider an educational system that has, I believe, a good chance for solving these major problems of learning. It involves a new paradigm for learning, via computer-based tutorial learning. Here I do not mean any of the expensive experiments involving artificial intelligence that are often identified with similar names. This is not a completely new system; we have been developing materials of this kind for over 30 years at the University of California, Irvine. Current technology, both computer power and bandwidth, is more than adequate, as we will see.

The following considerations are critical in the new learning system I propose:

• •

  Computers as tutors. The system is closely modeled after the learning situation involving a student and a human tutor. This implies an entirely different use of the computer than that typically seen today in learning.

• •

  Individualization of learning. Every student is different, in a variety of different ways. Each enters a given learning activity with her or his own interests, background, and learning skills. So the learning activity should be unique to each student, conscious of these differences.Students may require different times to learn something, so each moves at a unique pace. Decisions as what to present next to the learner must be taken very frequently, almost on a moment by moment basis. Many will learn faster than now. We want agile learning, as defined earlier.

• •

  Success in learning. Our goal should be that everyone succeeds in the learning activity, in all aspects. A failure to learn should be considered as a failure of the learning system, not of the student. The work of Bloom (1984) shows that this is possible with tutorial learning. It is unfair to students to settle for anything less.

• •

  Highly interactive learning. To accomplish individualization and success, the student needs to work in a highly interactive mode, just as she or he would with a human tutor. This is the "new" technology we require.

The typical interaction in our materials is a question from the computer, and a free-form answer (not multiple choice or pointing), all in the student's native language. In the future we believe the input will be primarily through voice. Current voice input systems are adequate to this task in the major languages, but not in many languages. Training of the computer for each student's voice is not required.

The frequency of interaction is also important. In experiments in public libraries with material designed for students of about 12 years of age, we found that if the interval between student inputs is greater than 20 seconds at any point in the dialog, we tend to lose some students. It is to be noted how different this is from current practices of lectures, video, and Web material. As we gain more experience with a range of students, we may find that this figure of 20 seconds may need to be adjusted for other types of students.

Note that the suggested interaction between the student and the computer resembles a conversation, or a Socratic dialog. It does not resemble a lecture. This becomes a question of design.

• •

  Learning with invisible tests. In a sense each question asked to the student is a test. The program is constantly exploring to find what the student needs to learn next. Since everyone will succeed, we are not trying to assign grades.So tests are invisible, non-existent, to the student. We need not worry about cheating. They will not experience the negative aspects of tests, often used to terrorize students. This will further encourage life-long learning, since learning is successful for each student.

• •

  Peer learning. Students will often work in groups of three or four, experiencing the benefits of learning from each other. These groups can either be working at the computer, connected remotely, or can be working at a task assigned by the computer or developed by the group.

• •

  Anywhere, anytime learning. These learning materials should be available and usable everywhere, whether it is in a schoolhouse in a wealthy country, or in remote villages, probably without schools, in a very poor part of the world.

Extensive development

Few materials of this type exist, so we need more information before proceeding to extensive development. We also need to organize the major development efforts necessary.

Technology

As suggested, existing technology is adequate to this new system. Current computers have more power and memory than is needed. Solar powered computers may be required in areas without electricity.

The materials may be delivered either through the Internet or, in areas without adequate telephones and Internet connections, with CD-ROM. Eventually we may deliver everywhere with two-way satellite connections.

Current Internet speeds, even slow telephone modems, will be adequate. Interactions will take place primarily at the user's computer, with segments of the program downloaded in the sense of Java applets. Streaming video will not be used, as it would violate the 20-second rule.

Speech recognition will need to be extended to other languages. Perhaps additional work will be needed to recognize the speech of very young children.

The production system for developing tutorial learning units exists, but should be improved before extensive development.

An agile learning system for all students is within our grasp. To better understand this, we need to consider how to produce such materials, and outline the steps for developing an educational system based on highly interactive tutorial learning. See the references for details.

AcknowledgementThis paper was dictated using Dragon NaturallySpeaking, version 5.

ReferencesAnnan, K. (1999), "Foreword", The State of the Word's Children, UNICEF, New York, NY.UNESCO (1996), Education for All, UNESCO, Jordon.

Further Reading Bork, Alfred, papers on this and related topics at www.ics.uci.edu/~borkBork, A. and Sigrun G. (2001), Tutorial Distance Learning: Rebuilding our Educational System, Kluwer, Boston, MA.Bloom, B. (1984), "The 2 sigma problem: the search for methods of group instruction as effective as one-to-one tutoring", Educational Researcher, July, pp. 4-15.

Alfred Bork(bork@uci.edu) is Professor Emeritus of Information & Computer Science at the University of California, Irvine, California, USA.