Example Based Pedagogical Strategies in a Computer Science Intelligent Tutoring System

Worked-out examples are a common teaching strategy that aids learners in understanding concepts by use of step-by-step instruction. Literature has shown that they can be extremely beneficial, with a large body of material showing they can provide benefits over regular problem solving alone. This research looks into the viability of using this teaching strategy in an intelligent tutoring system specificity designed for the computer science domain. Here, we detail the developed tutoring system, ChiQat-Tutor, which is designed with scalability and experimentation at its core. The system is described demonstrating its powerful architecture that gives the flexibility to analyse different teaching strategies. From the developed tutoring system, we focus on investigating the value in using the worked-out example teaching strategy in the system. Our investigation looks at human-human tutorial dialogues, and an implemented worked-out example module in ChiQat-Tutor for the linked list lesson. Multiple version of the worked-out example strategy is evaluated, with log data collected for each experimentation session that chronicles user behaviour. Various worked-out example based features are then identified that may be of used in such a system that could enhance student learning. We present a pipeline that may increase the effectiveness of an ITS that uses some of the example based features.