Recent development of teachable agent focuses on individualization and provides learners with active roles of knowledge constructors. The adaptive agent aims to maximize the learner’s cognitive functions as well as to enhance the learner’s interests in and motivation for learning. To establish the relationships between user characteristics and response patterns and to develop an algorithm based on the relationship, the individual characteristics of the learner were measured and the log data during interaction with the teachable agent named KORI were collected. A correlation analysis was conducted to identify the relationships among individual characteristics, user responses, and learning. Of the hundreds of possible relationships among numerous variables in three dimensions, nine key user responses were extracted, which were highly correlated with either the individual characteristics or learning outcomes. The results suggested that the construction of an individualized student model based on the ongoing response pattern of the user would be useful indices for predicting the learners’ individual characteristics and ongoing learning outcome. This study proposes a new type of method for assessing individual differences and dynamic cognitive/motivational changes rather than measuring them directly before or after learning.
[1]
Takashi Yamauchi,et al.
Learning from human tutoring
,
2001,
Cogn. Sci..
[2]
Gautam Biswas,et al.
Technology support for complex problem solving: from SAD environments to AI
,
2001
.
[3]
P. Pintrich,et al.
Motivation in Education: Theory, Research, and Applications
,
1995
.
[4]
Piet Kommers,et al.
Agent-support for problem solving through concept-mapping
,
2000
.