Background

Poor text comprehension in children of 7-11 years old is a matter of concern in schools that is recognised by teachers. Some children learn to decode accurately but then find comprehension difficult; they can read words but find it difficult to understand them. This tends to make reading less enjoyable which can reduce a child's motivation to read. This can significantly affect subsequent reading development.

Nicola Yuill has established a relationship between comprehension skill and language awareness (e.g. Yuill under review), assessed through children's understanding of riddles. It showed that poor comprehenders find it difficult to understand why jokes and riddles are funny. For example, in the joke:

the word spotted has a second, unexpected meaning. It is this that makes the joke funny. Arriving at this second interpretation is difficult for poor comprehenders. As a result, research investigated the link between riddle disambiguation and the development of comprehension skill. Software presented riddles to pairs of children and they were invited to discuss different interpretations in order to 'get the joke'. Pilot work showed that children's reading comprehension scores improved after interaction with the system.

Project Aims

The project team is multidisciplinary, drawing upon expertise in the fields of psychology (Nicola Yuill, Cindy Kerawalla and Amanda Harris), Human-Centred Technology (Rose Luckin) and computer science and linguistics (Darren Pearce). The overall aim of the Riddles project is to bring together ideas from all these fields to develop new software that can be used by pairs of children to improve their reading comprehension. The software will include many innovative features including:

• The presentation of jokes using not just text and static graphics but also animation and sound.
• Interaction with jokes will take place within a narrative to encourage more active engagement.
• Software scaffolding. Building upon previous work by Rose Luckin (e.g. Luckin 1999), the children's performance will be monitored and modelled by the system at all stages. This means that their progression through the narrative can be scaffolded and the level of help they receive tailored to their specific learning needs.
• Computer-mediated peer discussion. The system will also model and scaffold peer collaboration which is particularly challenging; as much as possible, the system must be able to derive salient information from collaboration between the children.