Wired and emotional program helps students discover the underlying meaning behind rules

AN emotionally responsive computer program called Clegg has boosted the debate on how to improve learning.

Bruce Geddes, a Durham University psychology postgraduate, and Rosemary Stevenson, reader in psychology at Durham and also a principal investigator at the human communication research centre at Edinburgh University, have been researching how students learn, using a word exchange game with a computer. And they conclude that while people are often unable to apply information learned in one situation to new problems, this can be overcome by looking for underlying patterns and meaning.

On the computer program, Clegg produces one of 12 emotional responses, ranging from "very loving'' to "cool'', depending on what word the player keys in, determined by a mathematical rule. One group of undergraduates was asked to make Clegg respond politely and stay polite, while the other group was asked to try to work out the pattern behind Clegg's responses. Both groups were then asked to make Clegg very friendly, and both did well, although the group searching for the pattern did better than those with the specific goal. They were then asked to predict Clegg's response to various words, with the pattern search group doing substantially better. This group also succeeded in describing the rule behind the game, which virtually none of the specific goal group could even attempt. "Most of the participants with the specific goal could only remember their specific responses, had no idea what the rule was and couldn't apply it,'' says Dr Stevenson.

She believes the research has implications for education in general, and educational software in particular.

"Computer assisted learning is going to stand or fall by how well you can get the learner to go and find the underlying pattern and rules,'' she says.

Another of her postgraduate students, Sharon MacDonald, has found that while pure hypertext with very little structure is unhelpful for retrieving a particular piece of information, the reverse is true in terms of learning. "We argue that's because with pure hypertext people are having to explore and discover for themselves, making their own pattern,'' she says.

Dr Stevenson believes that the learning process is similar for both adults and children, and that the new research could help the debate on "discovery'' learning versus rote learning in the classroom.

Discovery learning is often criticised, but Dr Stevenson suggests that what the children are supposed to be doing may not always be clear, with the danger that they go off track.

"It's important to have a learning goal. Children should be specifically directed to look for the patterns underlying what they are doing. They have to learn how to calculate areas, for example, and not just how to measure their playground."