Ph.D. in Computer Science thesis by Rania Hodhod. The University of York, Artificial Intelligence Group.
This thesis presents AEINS, Adaptive Educational Interactive Narrative System, that supports teaching ethics for 8-12 year old children. AEINS is designed based on Keller’s and Gagne’s learning theories. The idea is centered around involving students in moral dilemmas (called teaching moments) within which the Socratic Method is used as the teaching pedagogy. The important unique aspect of AEINS is that it exhibits the presence of four features shown to individually increase effectiveness of edugames environments, yet not integrated together in past research: a student model, a dynamic generated narrative, scripted branched narrative and evolving non-player characters. The student model aims to provide adaptation. The dynamic generated narrative forms a continuous story that glues the scripted teaching moments together. The evolving agents increase the realism and believability of the environment and perform a recognized pedagogical role by helping in supplying the educational process.
AEINS has been evaluated intrinsically and empirically according to the following themes: architecture and implementation, social aspects, and educational achievements. The intrinsic evaluation checked the implicit goals embodied by the design aspects and made a value judgment about these goals. In the empirical evaluation, twenty participants were assigned to use AEINS over a number of games. The evaluation showed positive results as the participants appreciated the social characteristics of the system as they were able to recognize the genuine social aspects and the realism represented in the game. Finally, the evaluation showed indications for developing new lines of thinking for some participants to the extent that some of them were ready to carry the experience forward to the real world. However, the evaluation also suggested possible improvements, such as the use of 3D interface and free text natural language.