The ability of CBT programs to adapt to the learning preferences of users could be transformed thanks to a new £1.25m EU-funded artificial intelligence project. By Alison Thomas
The trouble with human beings is that no two people or groups of people are the same. This can present problems in a training situation, although a sensitive and resourceful tutor will find ways of resolving them by changing the pace of instruction or adopting a different style of presentation.
Such versatility is generally absent in CBT, yet if it is to be educationally effective, it needs to become more sensitive to the needs of the user and find ways of catering for different preferred learning styles.
This was the thinking behind the launch in January 1998 of a £1.25m EU-funded project involving three organisations from the UK and three from Italy. The outcome is a highly adaptable multi-media program, which provides training on the European CENELEC wiring regulations. Although TEST (Training in Electrical Systems), has been designed to fulfil this very specific training need, the consortium that created it believes the resulting infrastructure could be used to imbue any piece of CBT with learning styles flexibility.
Start of a journey
“The TEST project is the start of a journey that will ultimately lead to more flexible, intelligent systems that deal with people on their own terms,” says Adrian Snook, a consultant with bespoke e-learning specialist VEGA Skillchange, one of the consortium members
“In the past, the best that courseware designers could hope for was to tailor materials to the middle ground. Adaptive training is less restrictive and allows you to reach people nearer the edges of the normal distribution curve.”
This flexibility is made possible by an artificial intelligence engine which makes deductions from interactions with the student and responds accordingly.
“If a student is not making progress along a pre-determined path, it will begin to shift the presentation mix,” he explains.
“It might start throwing in more video, more text or more graphics. It will then ask the question, ‘Have things improved?’ If the answer is no, it will try a different approach.”
The artificial intelligence was developed by Heriot Watt University in Edinburgh and the project has been managed by Giunti Multimedia of Milan.
Eastern Contracting and the Milan Electricity Generating company supplied subject matter expertise and Interactive Labs of Genoa contributed specialist resources and technology. The role of VEGA Skillchange has been to develop the training design and courseware components. Eighty training goals were identified and 80 training plans drawn up comprising a total of 800 components. Resources are re-usable and can be adapted to suit different contexts.
Collaborating across national boundaries has been a rewarding experience, although not everything has run smoothly. “We have had to contend with the bureaucracy of the European Commission and take account of the different training requirements across the European Union,” says Dr Keith Brown of the department of computing and electrical engineering at Heriot Watt University.
“The fact we have overcome these issues illustrates the strengths of the system. Because many of the resources are re-usable, you don’t have to reproduce all the material to meet different needs.”
After 27 months and a combined effort of 15.7 working years the moment of truth has arrived. Starting this month, two parallel groups of electricians will embark on training – one will follow a conventional tutor-led course, the other will use TEST.
Dr Brown is looking forward to the results. “We hope that they will highlight the advantages of adaptable training which is customisable to each individual in a self-paced learning environment,” he says. “Equally important to the university is the chance to identify any limitations so that we can restructure the system to overcome them.”
Only the beginning
The results of the trials will be made public in June, after which the programme will be made available commercially to organisations in the relevant sector. But that is only the beginning.
“The system has a lot of promise for anybody who is developing training materials for a population which is hard to define,” says Snook. “As with all technology, the challenge is finding the right application. I suspect the first adopters will be companies which haven’t got the luxury of being able to pre-select people with a learning style to suit their training programmes, which is often what happens.”
He does concede, however, that there are cost implications. “If, for example, you produce a different track to cater for visual, auditory and kinaesthetic preferences, that triples the content for the same training objective,” he explains. “So from the business point of view, there would have to be some kind of cost/benefit analysis. In situations where learning effectiveness is a key requirement, it will be worth the extra investment.”
In the short term, he believes his organisation has already benefited from participation in the project. “All of us have become much more aware of learning theory,” he says. “Even when we are building materials that are not adaptive, we are more conscientious about mixing the presentation in the optimum way to cater for learning styles.”
But his main interest now lies in the system’s potential for development.
“If there is anybody out there who’s interested in pursuing dual ventures using this technology for different ends, we are very keen to talk to them. This is a resource for the future.”
Crib sheet: learning styles
The most commonly used learning styles categorisation system is that proposed in the mid 1980s by Peter Honey and Alan Mumford. They identified four types of learner:
- Activists who enjoy doing things
- Reflectors who prefer quiet contemplation
- Theorists who like to understand the basic principles before applying them
- Pragmatists who shine at experimentation and practical problem-solving.
Research by Bandler, Grinder and Grinder draws attention to three further differences:
- Visual learners who respond best to visual stimuli
- Auditory learners who enjoy listening
- Kinaesthetic learners who prefer physical activity and interacting with others.
Most people combine several characteristics although one may predominate.
Preferred learning styles are also influenced by the fact that the left side of the brain controls logical thought while the right side is associated with creativity. Which side dominates will determine whether you are a “linear” or a “global” learner (Rose and Nicholl).