Design, Development and Evaluation of the Java Intelligent Tutoring System
Edward R. Sykes
The Java Intelligent Tutoring System (JITS) research project explored the power of a new approach to supporting beginner Java programming students. Using Java’s grammar as the core of its production rule base, JITS is embedded with extra functionality to detect, predict and correct lexicographical errors in students’ code. This additional functionality, developed during this research, is a plug-in to JITS called the Java Error Correction Algorithm (JECA).
One experimental group was provided the full JITS (with JECA); the second experimental group was given the skeleton JITS (without JECA); and the third was the control group which were taught in a traditional classroom environment. The findings showed that the group using the full version of JITS (with JECA) outperformed the skeleton JITS (without JECA) groups and the traditional classroom taught groups. There was a statistically significant difference (at the p = 0.05 level) between the JITS_JECA group and the JITS_ NOJECA group and between the JITS_JECA group and the control group.
Keywords: Programming Tutors, Intelligent Tutoring Systems, Designing Programming Tutors, Assessment of Programming Tutors.