Two different experiments on teaching how to program with active learning methodologies: a critical analysis

Abstract

To combat the difficulty that many students must learn how to program, the failure in introductory programming courses and the traditional high dropout rate, teachers have to use strategies that motivate and improve students' skills. Active methodologies and student-centered instruction can be a solution to get students interested on the subject, preparing assignments while learning in the classroom. This article reports on two very different experiences in two academic years. In the first year, agile SCRUM methodology, groups of five students, three interactions and a final project were used. In the second year, the Project based Learning was used with groups of three students for two different products, changing the composition of the groups. In both cases, peer classification was used. The results show that in the first case there is an increase in the approval rate, while in the second case there is an increase in the dropout rate. In this article we make a critical analysis of the results, analyzing what can be beneficial in one experiment and in the other in order to find an ideal model for using active methodologies to teach freshman computer science students how to program.