A study comparing test scores of a cooperative learning class and a traditional lecture-oriented class was conducted at a university in Japan. Two lecturers were assigned each class and teaching skill of these lecturers were confirmed to be practically equal from their previous performance.
For the cooperative learning class, LTD (Learning Through Discussion) with some modifications, revised LTD, was employed. In this class, students were grouped together in fives. They had to prepare the content of statistics and make a Mind Map (Buzan and Buzan, 1993) in advance. At the beginning of each class, students were given a mini-test of 15 to 20 minutes about the content of previous class. Because some of their evaluation was decided on the average mini-test scores of his/her group, the members of the group were interdependent. After the mini-test, using the Mind Map, students made a presentation of her/his preparation each other in his/her group.
For the traditional lecture-oriented class, a teacher lectured on statistics to all students each class, and students took notes.
At the end of the term, all students of both classes were tested about the content of statistics. This was the initial test. Three months after this initial test, the students took a second test, which was nearly identical to the previous one. Although the difference in initial test scores between the cooperative class and the traditional lecture-oriented class was not significant, comparing these scores with the scores of the second test did show a significant difference. The test scores of the students in the traditional lecture-oriented class declined significantly more than those of the students in the cooperative learning class. These results demonstrate that the students in the cooperative learning were able to retain more of the content of the class in a long time. Furthermore, these results indicate that ,for detecting the effectiveness of cooperative learning, researchers are required to evaluate it from longitudinal point of view.