Research in education recognizes the importance of matching the learning style of the student, strongly supporting the argument that a student learns best when the materials are presented in a form that matches her preferred learning style [1,2]. On this basis, some authors defend that effective lecturing should combine and balance the amount of cooperative, individualistic and competitive learning. Other authors are completely against the use of competitive strategies, arguing that this leads the student to regard others as obstacles to their own success, instead of as potential collaborators [3].

In this paper we present a case study in which the use of competitive and cooperative learning strategies are combined to allow knowledge transfer between a heterogeneous group of students with a diverse background knowledge on entrance to a course. In particular, we have used Robocode as part of a programming module in an attempt to encourage cooperation in a context in which prior student programming backgrounds were very diverse. As a first activity, students worked in teams and had to program the behaviour of a robot. These robots were then made to compete and the student achievement was measured as a function of the number of opponents they were able to beat. As a second activity, students had to individually improve their robots. This time, student achievement was measured as a function of the improvement with respect to their original implementation. During this second activity, students were allowed to help each other, in a peer to peer fashion. When this happened, the student with the best results in the first activity takes the role of an advisor and the increase in grade due to the improvement achieved is shared between both members. The final mark for the assessment was the addition of the marks obtained in both rounds. This grading system motivates students with the best results to collaborate with those who implemented the worst robots.

From a collaborative perspective, during the first activity students worked in small groups to achieve a common objective: designing a better strategy than other groups. Although this is in nature a competitive strategy, we also allowed students to obtain the minimum passing grade on a non-competitive basis. The basic requirement to obtain this grade was to design a robot which was able to beat some already implemented robots provided with the environment, an activity which did not depend on the achievements of others. In the second activity, students worked in pairs. In this occasion, the grades obtained were dependent upon the improvement achieved, encouraging the best students to help those who obtained the worst results in the previous activity. Overall, this strategy fostered collaboration between students with different background knowledge, facilitating knowledge transfer between peers.


[1] Johnson, D. W. and Johnson, R. T., “Learning together and alone: cooperative, competitive and individualistic learning (5th edition)”, 1998, Needham Heights, MA: Allyn & Bacon

[2] Butler, R., “Interest in the task and interest in peer’s work in competitive and non-competitive conditions; a developmental study”, Child development 60(3), pp. 562-570, 1989

[3] Kohn, A. “No Contest: The Case Against Competition”, Boston: Houghton Mifflin, 1986.