Development and Evaluation of a Web-based Diagnostic System for Beginners Programming Course

Abstract

Beginners’ programming courses have different learners with different learning processes and understandings. Teachers of these classes cannot easily monitor and assess the learners’ learning levels and give them appropriate advices in the limited class time. Thus these courses need a diagnostic system, which can easily assess learners’ understanding and give them advices. This study firstly analyzed the past programming test results with decision tree and IRS (Item Relational Structure) analyses, and made a learners’ classification model. Based on this model, we developed a Web-based diagnostic system for beginners programming course. The system gives a few questions to learners in decision tree order, shows the diagnostic result and gives advices promptly. Finally this study evaluates the Web-based system on its usability and fitness of the classification model. The result shows validity of the model and high usability of the system.