The Effects of Peer Tutoring and Teachable Agent on Interest and Task Performance

The Effects of Peer Tutoring and Teachable Agent on Interest and Task Performance Yeonhee So (yukgam91@kyungnam.ac.kr) Department of Education, Kyungnam University, Masan, 631-701, South Korea Youngok Kim (youokim@ kyungnam.ac.kr) Department of Mathematical Education, Kyungnam University, Masan, 631-701, South Korea due to the tutors’ inexperience in teaching skills (Kim et al., 2003). Peer tutoring also has restrictions in space and time, while unnecessary interactions between tutor and tutee, which can interfere with the learning process, might occur with younger participants. Various highly developed fields of information and technology are presently available. Accordingly, with the development of computers and communication technologies, students are growing up with technology. It is therefore desirable to effectively utilize such technologies in education. In fact, the traditional computer assisted learning (CAL) system has been utilized in educational settings for a long time. However, CAL systems such as the intelligent tutoring system (ITS) are based on passive learning activities in which the students are provided with learning materials and required to memorize them repeatedly via CAL. Thus, many researches have criticized the iterative and passive practice problems of CAL. In addition, traditional CAL does not reflect individual differences such as learner’s cognitive ability and motivational aspects. Actually, the use of an identical interface, regardless of the individual differences, might be not only less effective in cognitive aspects of learning but also less interesting in terms of motivation. To overcome the limitations of peer tutoring and CAL, Schwartz et al (2000) proposed the new concept of learning by teaching through an agent called the Teachable Agent (TA). TA is a computer program in which students teach the computer agent to enhance their motivation and cognitive ability based on the instructional method of ‘learning by teaching’. That is, a computer-based system utilizing the benefits gained from the act of teaching has been developed to use TA as one of the ITS programs. In this TA, the agent provides student tutors with an active role and positive attitude toward the subject matter. Thus, learners are enabled to organize and acquire problem solving knowledge about various domains for instructing an intelligent agent (Biswas et al., 2001) In this study, we developed a kind of TA, the KORea university intelligent agent (KORI) (Kim et al., 2005a, 2005b), and investigated its effectiveness in comparison with peer tutoring on students’ task performance and interest according to their level of self-efficacy. Abstract This experiment was conducted to compare the effects of peer tutoring with an intelligent teachable agent (TA), the Korea university intelligent agent (KORI), on students’ interest and task performance according to their level of self-efficacy. The results showed a significant interaction effect on interest between the peer tutoring/KORI and the level of self-efficacy. The high self- efficacious group had greater interest in peer tutoring than in KORI, whereas the low self-efficacious group had greater interest in KORI than in peer tutoring. Analysis of the task performance revealed that there was a main effect of peer tutoring/KORI and interaction effect on the task performance between peer tutoring/ KORI and the level of self-efficacy. The participants with high self-efficacy received high scores in both peer tutoring and KORI, whereas the participants with low self-efficacy gained higher scores in KORI than in peer tutoring. Keywords: teachable performance agent, peer tutoring, interest, task Introduction Peer tutoring is an effective learning method based on the concept of learning by teaching. Previous studies have provided plenty of evidence that peer tutoring is an effective method of learning for both tutor and tutee (e.g., Kulik & Kulik, 1982). Ginsbug-Block and Fantuzzo (1997) reported that peer tutoring enhanced the tutee’s academic achievement, social relationship, self-concept and motivation to learn, while simultaneously promoting the tutor’s patience and ability for task performance, self- control, and motivation. Furthermore, peer tutoring increased the tutor’s positive attitude toward the tutee and basic understanding of the subject areas (Cohen et al., 1982). Thus, peer-tutoring activities have been regarded as a meaningful learning method for improving cognitive ability and academic motivation for both tutor and tutee. Despite these potential benefits, peer tutoring has some limitations in practical learning settings. In face-to-face peer tutoring, tutors can experience a cognitive burden because of the large amount of information they are required to remember for effective teaching, and thus lose confidence in tutoring. In addition, we can’t completely rule out the possibility that tutees don’t perfectly understand what tutors teach and even worse may learn misconceptions because