Modeling of Programming Learning Behavior with Contextual and Progressive Data (教育工学)

概要

論文の詳細を見る
All programming educators want to design a course that will keep all of their students making efforts on the learning throughout the course, taking full advantage of their training condition. However, it is very hard to achieve such a good course with present design methods. If they apply training strategies such as pair programming or game programming same to all students groups, some groups who cannot adapt to these strategies may not get along well with the course. If they use different strategies for different students, they need to know what strategies are good for what kinds of the students. This paper proposes a method to design a programming course, using motivation and cost correlation models. The models present the motivation of a specific student group under an amount of cost of a specific strategy. They also show the correlation of motivation and cost of the course plan on student groups along the semester. The models are derived from student genuine learning behaviors taken with a programming support web-based system. Good correlations will be taken into for the refinement. We have succeeded in refining a course plan, applying the method to successive 2 programming courses, each of which consists of approximately 500 students.
一般社団法人電子情報通信学会の論文
2011-10-28