Adaptivity in Problem-based Learning: Use of Granularity Sally He, Massey University, New Zealand Kinshuk, Massey University, New Zealand Hong Hong, Massey University, New Zealand Ashok Patel, De Montfort University, United Kingdom Abstract This paper discusses the limitations of PBL learning environments and introduces the student adaptivity technology into PBL environments to improve the effectiveness and efficiency of the learning process. A web- based prototype is implemented by using PHP, MySQL and Apache, and uses the accounting as subject domain. With the system, students work on the real world costing calculation problems, and the system evaluates students’ performance results on the problems to provide adaptation to the students. 1. Introduction Constructivist approach to learning has been around for quite some time. The constructivist theory has resulted in the development of a wide variety of learning environments, however the problem-based learning (PBL) lends itself as one of the most suitable candidates for its deployment [2]. PBL is an attractive approach to foster learner’s critical problem solving and self-directed learning skills. However, it is difficult to implement effective PBL environments. A majority of existing PBL environments suffers from the fact that the students easily get inundated by the fine granularity of the problems and loose focus of overall aims of the learning process. This paper describes a problem-based learning environment that attempts to address this problem by introducing student adaptivity technologies. 2. Constructivism Theory and Problem-based Learning Like other learning theories, constructivism has multiple roots in the philosophical and psychological viewpoints of this century. A number of contemporary cognitive theorists have adopted constructivism theory, which considers: knowledge is a function of how the individual creates meaning from his/her own experiences during learning and understanding [1]. This theory is characterized by following three propositions [2]. A) Knowledge is in the interaction of humans with the environment (the core concept of constructivism). B) Cognitive conflict is the stimulus for learning and determines the organization and nature of what is learned: when human beings are in a learning environment, there is some stimulus for learning. C) Understanding is influenced by the processes associated with collaborative learning. The features of constructivism outlined above have been the basis of a wide variety of learning environments, including problem-based learning environments. The problem-based learning model has its roots in the apprenticeship. It emphasizes a "real-world" approach to learning: a student-centered process that is both constructive and collaborative. PBL is a motivating way to learn because learners are involved in active learning, working with real problems. Within PBL environment, students are able to build and improve their problem- solving and self-directed learning skills. However, in practice, PBL is difficult to implement, with or without computer-based support. In the traditional face- to-face PBL, teachers must be specially trained as guides and students often become frustrated by the lack of information. In the computer assisted intelligent PBL environment, since the PBL does not limit what students may choose to learn, and the process may provide little guidance on the best ways of achieving learning goals, students may be concerned that their learning strategies are misdirected or inefficient. Thus, it is much harder for student in the computer intelligent learning systems with PBL, and students easily get lost during learning and become frustrated by feeling out of control in their study. 3. Student Adaptivity in Computer Intelligent Learning Systems Student adaptivity in intelligent learning systems provides the systems ability to adapt themselves to the goals and tasks of student by monitoring their performance. The adaptivity is one of the core components in intelligent learning environment. The main reasons that student adaptivity is so important to intelligent learning systems are as follows: • A wide student spectrum: the student spectrum may be from one extreme (naïve) to another extreme (advanced), that means that students may have Proceedings of the International Conference on Computers in Education (ICCE’02) 0-7695-1509-6/02 $17.00 © 2002 IEEE