Abstract:
Although several tools for computerized adaptive practicing are available, implementing adaptive practicing with generators remains complex. This study proposes a framework for computerized adaptive practice (CAP) using the ELO rating system. The framework aims to help convert question generators to a CAP system as a web-service application programming interface, which can be easily integrated with applications. The framework comprises four models (indicator, learner, question instance, and script) and seven customizable modules (question instance selector, question instance generator, distance calculator, question display, answer checker, feedback display, and question instance updater). With the separated question instance generator module and question display module, the framework supports question domains (e.g., mathematics, science, and language) in any form; the question bank can be scaled automatically when there is no question that matches a learners preference. This feature fills the gap in the literature regarding the existing tools, where educators are required to devise questions, leading to difficulty in scaling for various sizes of learner groups. Five types of reports can be used for analytics. An adaptive practicing system can support online and offline classrooms of any size. Moreover, learners can practice at their pace; educators are not required to handcraft questions for the system.