Association for the Advancement of Computing in Education
This paper describes how to dynamically monitor each student's learning performance using causal modeling approach in computer networked learning environments. Although using the computer network technology to deliver instruction has become possible and has resulted in many advantages, most of the current implementation seems ineffective in gaining a better understanding about each learner's detailed interaction behavior. Only when each learner's performance can be more accurately captured, can the instruction system better adapt a subsequent pedagogical strategy for the learners. In order to promote an effective distributed learning environment, a system must discern each learner's dialog behavior pattern as well as more accurately determine the learner's performance level. Other considerations include a system's automatic verification of requests for human assistance and prioritizing the sequence of help when situations require. This paper investigates the nature of the dialog interaction especially in the network based environments and proposes an appropriate methodology that can concurrently capture all learners' performance along the interactive session.
International Journal of Educational Telecommunications 1(2), pp.309-321