THE EFFECT OF A ONE-ON-ONE DIALOGUE-BASED MATHEMATICAL INTELLIGENT TUTORING SYSTEM FOR LEARNING EQUIVALENT FRACTION

Abstract

The purposes of this study are to develop a one-on-one dialogue-based mathematical intelligent tutoring system (ITS) for learning equivalent fraction in the 4th grade math, and evaluate its learning effect. The system used the course content and dialogue script designed by the math teaching experts in advance, and a computer agent teacher asked questions based on the course script. After the student answered, the system was able to identify the error pattern and misconception according to the student's response, then provided each student with adaptive teaching guidance or feedback. Students could construct correct equivalent fraction concepts through a series of interactive dialogues between students and the computer agent teacher step by step. In order to evaluate the effectiveness of this ITS, a quasi-experiment design was conducted. The pretest and post-test were parallel tests involving equivalent fraction. The participants of the study were 76 students in the fourth grade of two elementary schools chosen from midland of Taiwan. They were divided into the experimental group of 39 and the control group of 37. The experimental group used the "one-on-one dialogue-based mathematical intelligent tutoring system" for teaching. The control group used traditional classroom instruction by a human teacher. The learning content and time were controlled to be the same. Finally, the learning effectiveness and learning interest were assessed by comparing the pre-test and post-test performance of students. The results of the study showed that both teaching methods can significantly improve the students’ learning achievements of equivalent fraction, and the learning effectiveness of "one-on-one dialogue-based mathematical intelligent tutoring system" was significantly better than that of traditional classroom instruction. In the "one-on-one dialogue-based mathematical intelligent tutoring system" group, the learning improvement of students with different genders and different ability levels were also reaching a significant level. It indicated that this system benefited the learning achievements of students with different genders and different abilities. Furthermore, from the response data of the learning interest questionnaire, both teaching methods could significantly improve the learning interest of students. But there was no significant difference between the two teaching methods. By interviewing students, the probable causes included that low learning interest students of the experimental group also lack interest to familiar system operation, and some students think this ITS is not interesting enough because of lacking learning games.